Upload folder using huggingface_hub

annotation_data/sample_data/sample_pdf_links.json

@@ -9,5 +9,126 @@
     "has_revalidation": true,
     "has_datasets": true,
     "num_pages": 12
+  },
+  {
+    "index": 2,
+    "src_docname": "doc_2",
+    "original_name": "2025 03  SEIS ukraine_socio-economic_researchpaper",
+    "landing_page_url": "/sample_pdfs/2025 03  SEIS ukraine_socio-economic_researchpaper.pdf",
+    "direct_pdf_url": "/sample_pdfs/2025 03  SEIS ukraine_socio-economic_researchpaper.pdf",
+    "status": "success",
+    "has_revalidation": true,
+    "has_datasets": true,
+    "num_pages": 16
+  },
+  {
+    "index": 3,
+    "src_docname": "doc_3",
+    "original_name": "Analysis of the impact of refugees from Ukraine on the economy of Poland",
+    "landing_page_url": "/sample_pdfs/Analysis of the impact of refugees from Ukraine on the economy of Poland.pdf",
+    "direct_pdf_url": "/sample_pdfs/Analysis of the impact of refugees from Ukraine on the economy of Poland.pdf",
+    "status": "success",
+    "has_revalidation": true,
+    "has_datasets": true,
+    "num_pages": 25
+  },
+  {
+    "index": 4,
+    "src_docname": "doc_4",
+    "original_name": "LCRP_2021",
+    "landing_page_url": "/sample_pdfs/LCRP_2021.pdf",
+    "direct_pdf_url": "/sample_pdfs/LCRP_2021.pdf",
+    "status": "success",
+    "has_revalidation": true,
+    "has_datasets": true,
+    "num_pages": 212
+  },
+  {
+    "index": 5,
+    "src_docname": "doc_5",
+    "original_name": "LCRP_2025",
+    "landing_page_url": "/sample_pdfs/LCRP_2025.pdf",
+    "direct_pdf_url": "/sample_pdfs/LCRP_2025.pdf",
+    "status": "success",
+    "has_revalidation": true,
+    "has_datasets": true,
+    "num_pages": 192
+  },
+  {
+    "index": 6,
+    "src_docname": "doc_6",
+    "original_name": "LebanonResp2025",
+    "landing_page_url": "/sample_pdfs/LebanonResp2025.pdf",
+    "direct_pdf_url": "/sample_pdfs/LebanonResp2025.pdf",
+    "status": "success",
+    "has_revalidation": true,
+    "has_datasets": true,
+    "num_pages": 192
+  },
+  {
+    "index": 7,
+    "src_docname": "doc_7",
+    "original_name": "NAVIGATING HEALTH AND WELL-BEING CHALLENGES FOR REFUGEES FROM UKRAINE - 2nd Edition",
+    "landing_page_url": "/sample_pdfs/NAVIGATING HEALTH AND WELL-BEING CHALLENGES FOR REFUGEES FROM UKRAINE - 2nd Edition.pdf",
+    "direct_pdf_url": "/sample_pdfs/NAVIGATING HEALTH AND WELL-BEING CHALLENGES FOR REFUGEES FROM UKRAINE - 2nd Edition.pdf",
+    "status": "success",
+    "has_revalidation": true,
+    "has_datasets": true,
+    "num_pages": 32
+  },
+  {
+    "index": 8,
+    "src_docname": "doc_8",
+    "original_name": "Poland Analysis of the impact of refugees from Ukraine on the economy of Poland – March 2024",
+    "landing_page_url": "/sample_pdfs/Poland Analysis of the impact of refugees from Ukraine on the economy of Poland – March 2024.pdf",
+    "direct_pdf_url": "/sample_pdfs/Poland Analysis of the impact of refugees from Ukraine on the economy of Poland – March 2024.pdf",
+    "status": "success",
+    "has_revalidation": true,
+    "has_datasets": true,
+    "num_pages": 24
+  },
+  {
+    "index": 9,
+    "src_docname": "doc_9",
+    "original_name": "Revised Ukraine Situation 2025 2026 RRP FINAL (5)",
+    "landing_page_url": "/sample_pdfs/Revised Ukraine Situation 2025 2026 RRP FINAL (5).pdf",
+    "direct_pdf_url": "/sample_pdfs/Revised Ukraine Situation 2025 2026 RRP FINAL (5).pdf",
+    "status": "success",
+    "has_revalidation": true,
+    "has_datasets": true,
+    "num_pages": 183
+  },
+  {
+    "index": 10,
+    "src_docname": "doc_10",
+    "original_name": "SESRE_WBG",
+    "landing_page_url": "/sample_pdfs/SESRE_WBG.pdf",
+    "direct_pdf_url": "/sample_pdfs/SESRE_WBG.pdf",
+    "status": "success",
+    "has_revalidation": true,
+    "has_datasets": true,
+    "num_pages": 146
+  },
+  {
+    "index": 11,
+    "src_docname": "doc_11",
+    "original_name": "Somalia_Baseline",
+    "landing_page_url": "/sample_pdfs/Somalia_Baseline.pdf",
+    "direct_pdf_url": "/sample_pdfs/Somalia_Baseline.pdf",
+    "status": "success",
+    "has_revalidation": true,
+    "has_datasets": true,
+    "num_pages": 17
+  },
+  {
+    "index": 12,
+    "src_docname": "doc_12",
+    "original_name": "Somalia_NSP",
+    "landing_page_url": "/sample_pdfs/Somalia_NSP.pdf",
+    "direct_pdf_url": "/sample_pdfs/Somalia_NSP.pdf",
+    "status": "success",
+    "has_revalidation": true,
+    "has_datasets": true,
+    "num_pages": 7
   }
 ]

annotation_data/sample_extractions/doc_1/raw/doc_1_direct_judged.jsonl

@@ -1,7 +1,63 @@
 [
   {
     "input_text": "HELPING HANDS\nTHE ROLE OF \nHOUSING \nSUPPORT AND \nEMPLOYMENT \nFACILITATION IN \nECONOMIC \nVULNERABILITY \nOF REFUGEES \nFROM UKRAINE\nAn inter-agency \nexploration of socio-\neconomic data\nApril 2024\n",
-    "datasets": [],
+    "datasets": [
+      {
+        "dataset_name": {
+          "text": "HELPING HANDS THE R",
+          "confidence": 1,
+          "start": null,
+          "end": null
+        },
+        "dataset_tag": "named",
+        "source": "human",
+        "annotator": "rafmacalaba",
+        "timestamp": "2026-06-07T23:18:06.443Z",
+        "description": null,
+        "data_type": null,
+        "acronym": null,
+        "author": null,
+        "producer": null,
+        "geography": null,
+        "publication_year": null,
+        "reference_year": null,
+        "reference_population": null,
+        "is_used": null,
+        "usage_context": null
+      },
+      {
+        "dataset_name": {
+          "text": "HOUSING SUPPORT AND EMPLOYMENT FACILITATION IN",
+          "confidence": 1,
+          "start": 27,
+          "end": 76
+        },
+        "dataset_tag": "named",
+        "source": "human",
+        "annotator": "rafmacalaba",
+        "timestamp": "2026-06-07T23:22:23.382Z",
+        "description": null,
+        "data_type": null,
+        "acronym": null,
+        "author": null,
+        "producer": null,
+        "geography": null,
+        "publication_year": null,
+        "reference_year": null,
+        "reference_population": null,
+        "is_used": null,
+        "usage_context": null,
+        "validations": [
+          {
+            "human_validated": true,
+            "human_verdict": true,
+            "human_notes": null,
+            "annotator": "rafmacalaba",
+            "validated_at": "2026-06-07T23:23:05.736Z"
+          }
+        ]
+      }
+    ],
     "document": {
       "source": "/sample_pdfs/2024 04 17 Helping Hands - An inter-agency exploration of socio economic data Ukraine final_3.pdf",
       "pages": [
@@ -708,52 +764,6 @@
         "is_used": "True",
         "usage_context": "supporting",
         "validations": []
-      },
-      {
-        "dataset_name": {
-          "text": "(SILC) survey",
-          "confidence": 1,
-          "start": null,
-          "end": null
-        },
-        "dataset_tag": "vague",
-        "source": "human",
-        "annotator": "rafmacalaba",
-        "timestamp": "2026-06-07T22:50:45.395Z",
-        "description": null,
-        "data_type": null,
-        "acronym": null,
-        "author": null,
-        "producer": null,
-        "geography": null,
-        "publication_year": null,
-        "reference_year": null,
-        "reference_population": null,
-        "is_used": null,
-        "usage_context": null
-      },
-      {
-        "dataset_name": {
-          "text": "(SILC) survey",
-          "confidence": 1,
-          "start": 2917,
-          "end": 2931
-        },
-        "dataset_tag": "vague",
-        "source": "human",
-        "annotator": "rafmacalaba",
-        "timestamp": "2026-06-08T01:24:07.017Z",
-        "description": null,
-        "data_type": null,
-        "acronym": null,
-        "author": null,
-        "producer": null,
-        "geography": null,
-        "publication_year": null,
-        "reference_year": null,
-        "reference_population": null,
-        "is_used": null,
-        "usage_context": null
       }
     ],
     "document": {
@@ -855,29 +865,6 @@
         "is_used": "True",
         "usage_context": "primary",
         "validations": []
-      },
-      {
-        "dataset_name": {
-          "text": "(SILC) survey",
-          "confidence": 1,
-          "start": null,
-          "end": null
-        },
-        "dataset_tag": "vague",
-        "source": "human",
-        "annotator": "rafmacalaba",
-        "timestamp": "2026-06-07T22:50:09.820Z",
-        "description": null,
-        "data_type": null,
-        "acronym": null,
-        "author": null,
-        "producer": null,
-        "geography": null,
-        "publication_year": null,
-        "reference_year": null,
-        "reference_population": null,
-        "is_used": null,
-        "usage_context": null
       }
     ],
     "document": {

annotation_data/sample_extractions/doc_11/raw/doc_11_direct_judged.jsonl

@@ -0,0 +1,445 @@
+[
+  {
+    "input_text": "Power BI Desktop\nSolutions \nPathways\nAction Plan:\nBaseline Analysis\n2024 - 2029\nTHE FEDERAL GOVERNMENT OF \nSOMALIA MINISTRY OF PLANNING \nINVESTMENT AND ONOMIC \nDEVELOPMENT\n",
+    "datasets": [],
+    "document": {
+      "source": "/sample_pdfs/Somalia_Baseline.pdf",
+      "pages": [
+        0
+      ]
+    }
+  },
+  {
+    "input_text": "Power BI Desktop\nSomalia’s Solutions \nPathways\nThe Department of Poverty Reduction and Durable Solutions of the \nMinistry of Planning, Investments, and Economic Development \n(MoPIED) in Somalia continues to play a leading role in overseeing \nand addressing solutions to internal displacement in the country. \nIn 2024, the department conducted a series of durable solutions \npathways consultations across all federal member states and the \nBanadir Regional Administration (BRA), resulting in the National \nSolutions Pathways Action Plan. These consultations aimed to \nidentify key approaches to facilitate the durable solutions \npathways of one million Internally Displaced Persons (IDPs) from \ntheir displacement situations, in alignment with the National \nDurable Solutions Strategy, the National Development Plan and the \nUnited Nations Secretary-General’s Action Agenda on Internal \nDisplacement, emphasising three interconnected goals: achieving \ndurable solutions, ensuring protection and assistance, and \npreventing future displacement. \nBaseline Analysis\nThis document presents key quantitative data under each of the \nsix Solutions Pathways: 1) Government leadership, 2) Access to \nBasic Services, 3) Employment and Livelihood Opportunities, 4) \nLegal Documentation, HLP and Access to Justice, 5) Climate Change \nand Resilience. Pathway 6 refers to Data for Solutions, which is \ncross-cutting and covers MoPIED's Durable Solutions Progress \n(DSP) Survey, was supported by IOM Somalia in 2024. Data from \nthis survey is presented under each of the six pathways in this \ndocument. Data collected by UNHCR related to protection and \nconflict sensitivity is also presented, highlighting key cross-cutting \nissues to consider in implementing the plan.\n",
+    "datasets": [
+      {
+        "dataset_name": {
+          "text": "National Solutions Pathways Action Plan",
+          "confidence": 0.8032320141792297,
+          "start": 488,
+          "end": 527,
+          "clean_text": "National Solutions Pathways Action Plan"
+        },
+        "dataset_tag": "named",
+        "description": null,
+        "data_type": "other",
+        "acronym": {
+          "text": "",
+          "confidence": 0.0,
+          "start": 488,
+          "end": 527,
+          "clean_text": ""
+        },
+        "author": null,
+        "producer": {
+          "text": "",
+          "confidence": 0.0,
+          "start": 488,
+          "end": 527,
+          "clean_text": ""
+        },
+        "geography": {
+          "text": "Somalia",
+          "confidence": 0.9999734163284302,
+          "start": 17,
+          "end": 24,
+          "clean_text": "Somalia"
+        },
+        "publication_year": null,
+        "reference_year": {
+          "text": "",
+          "confidence": 0.0,
+          "start": 488,
+          "end": 527
+        },
+        "reference_population": null,
+        "is_used": "True",
+        "usage_context": "primary",
+        "validations": []
+      },
+      {
+        "dataset_name": {
+          "text": "Durable Solutions Progress (DSP) Survey",
+          "confidence": 0.4085761606693268,
+          "start": 1399,
+          "end": 1438,
+          "clean_text": "Durable Solutions Progress (DSP) Survey"
+        },
+        "dataset_tag": "named",
+        "description": null,
+        "data_type": "other",
+        "acronym": {
+          "text": "",
+          "confidence": 0.0,
+          "start": 1399,
+          "end": 1438,
+          "clean_text": ""
+        },
+        "author": null,
+        "producer": {
+          "text": "",
+          "confidence": 0.0,
+          "start": 1399,
+          "end": 1438,
+          "clean_text": ""
+        },
+        "geography": {
+          "text": "",
+          "confidence": 0.0,
+          "start": 1399,
+          "end": 1438,
+          "clean_text": ""
+        },
+        "publication_year": null,
+        "reference_year": {
+          "text": "2024",
+          "confidence": 0.9948785305023193,
+          "start": 1472,
+          "end": 1476
+        },
+        "reference_population": null,
+        "is_used": "True",
+        "usage_context": "primary",
+        "validations": []
+      },
+      {
+        "dataset_name": {
+          "text": "Data collected by UNHCR",
+          "confidence": 0.7616613507270813,
+          "start": 1562,
+          "end": 1585,
+          "clean_text": "Data collected by UNHCR"
+        },
+        "dataset_tag": "named",
+        "description": null,
+        "data_type": "survey",
+        "acronym": {
+          "text": "",
+          "confidence": 0.0,
+          "start": 1562,
+          "end": 1585,
+          "clean_text": ""
+        },
+        "author": null,
+        "producer": {
+          "text": "UNHCR",
+          "confidence": 0.7918815612792969,
+          "start": 1580,
+          "end": 1585,
+          "clean_text": "UNHCR"
+        },
+        "geography": {
+          "text": "Somalia",
+          "confidence": 0.999549925327301,
+          "start": 1461,
+          "end": 1468,
+          "clean_text": "Somalia"
+        },
+        "publication_year": null,
+        "reference_year": {
+          "text": "",
+          "confidence": 0.0,
+          "start": 1562,
+          "end": 1585
+        },
+        "reference_population": null,
+        "is_used": "True",
+        "usage_context": "supporting",
+        "validations": []
+      }
+    ],
+    "document": {
+      "source": "/sample_pdfs/Somalia_Baseline.pdf",
+      "pages": [
+        1
+      ]
+    }
+  },
+  {
+    "input_text": "Power BI Desktop\nSolutions Pathways Action Plan\nLinking the Frameworks (NDSS >> Pathways >> IASC)\nOverview: Solutions Policy Frameworks in Somalia\n",
+    "datasets": [],
+    "document": {
+      "source": "/sample_pdfs/Somalia_Baseline.pdf",
+      "pages": [
+        2
+      ]
+    }
+  },
+  {
+    "input_text": "Power BI Desktop\nState\nIDP Individuals\nHC Individuals\nTargeted Individuals\nin Action Plan\nBanadir\n920,532\n95,928\n299,360\nDayniile\n390,447\n426\n156,896\nKahda\n530,085\n95,502\n142,464\nGalmudug\n147,517\n479,788\n65,808\nCabudwaaq\n48,039\n142,164\n21,060\nCadaado\n14,790\n200,436\n11,539\nGaalkacyo\n69,348\n137,188\n17,869\nGuriel\n15,340\n15,340\nHirshabelle\n156,529\n1,385,431\n31,925\nBalcad\n21,224\n462,884\n8,576\nBelet Weyne\n100,561\n286,129\n13,742\nBulo Burto\n1,080\n146,709\n482\nJowhar\n33,664\n489,709\n9,125\nJubaland\n343,032\n352,776\n151,219\nDhobley\n49,536\n49,536\nDoolow\n124,062\n37,458\n41,276\nKismaayo\n169,434\n315,318\n60,407\nPuntland\n220,312\n592,092\n69,885\nBossaso\n99,880\n248,336\n32,590\nGaalkacyo\n49,096\n104,346\n17,869\nGaroowe\n54,656\n125,670\n14,148\nQardho\n16,680\n113,740\n5,278\nSomaliland\n275,799\n901,919\n131,555\nBurco\n65,610\n96,029\n21,222\nBuuhoodle\n870\n15,000\n8,200\nCaynabo\n12,900\n67,044\n4,457\nCeerigaabo\n10,170\n38,010\n9,667\nHargeysa\n111,720\n681,756\n64,689\nLaas Caanood\n74,529\n18,200\nTaleex\n0\n1,320\n3,111\nXudun\n0\n2,760\n2,009\nSouth West\n854,504\n989,470\n237,559\nAfgooye\n57,084\n283,224\n16,451\nBaidoa\n731,492\n435,715\n185,894\nDiinsoor\n12,180\n109,498\n26,461\nXudur\n53,748\n161,033\n8,753\nTotal\n2,918,225\n4,797,404\n987,311\nThe below information presents population estimates in each of the priority districts included in the \nSolutions Pathways Action Plan. Refer to the DTM Baseline (Round 3) database here: \nhttps://dtm.iom.int/online-interactive-resources/somalia-baseline-assessment-2024-dashboard\nBanadir state\n920,532\nHC 95,928\nIDP\nGalmadug state\n147,517\n479,788\nIDP\nHC\nHirshabelle state\n156,529\n1,385,431\nIDP\nHC\nJubaland state\n343,032\n352,776\nIDP\nHC\nPuntland state\n220,312\n592,092\nIDP\nHC\nSomaliland state\n275,799\n901,919\nIDP\nHC\nSouth-west state\n854,504\n989,470\nIDP\nHC\nSolutions Pathways Action Plan\nPopulation Overview\nSelected Districts - Solutions Pathways Action Plan\n",
+    "datasets": [],
+    "document": {
+      "source": "/sample_pdfs/Somalia_Baseline.pdf",
+      "pages": [
+        3
+      ]
+    }
+  },
+  {
+    "input_text": "Power BI Desktop\nGovernment \nLeadership\nEnsuring a whole of governmental \napproach \nwith \noversight \nand \ninvolvement \nis \ncrucial \nfor \nsuccessfully \nidentifying \nand \nimplementing solutions. Effective \nleadership and coordination at the \nnational and local levels are \nnecessary to ensure systematic \ninclusion of displacement-affected \ncommunities \n(DACs) \nacross \ngovernment \npriorities \nand \nplanning processes, as well as \nsuccessfully implementing durable \nsolutions programs.\nPathway #1:\n",
+    "datasets": [],
+    "document": {
+      "source": "/sample_pdfs/Somalia_Baseline.pdf",
+      "pages": [
+        4
+      ]
+    }
+  },
+  {
+    "input_text": "Power BI Desktop\nLocal government inclusiveness\nBaardheere\nBaydhaba\nBerdaale\nDoolow\nKismaayo\nMogDaynile\nMogKahda\nXudur\n65%\n77%\n84%\n83%\n81%\n71%\n71%\n76%\n80%\n82%\n91%\n84%\n85%\n53%\n63%\n73%\nLocal government responsive to community needs\nBaardheere\nBaydhaba\nBerdaale\nDoolow\nKismaayo\nMogDaynile\nMogKahda\nXudur\n67%\n81%\n82%\n84%\n79%\n64%\n64%\n70%\n77%\n83%\n90%\n81%\n85%\n46%\n61%\n76%\nRegular attendance in public meetings\nBaardheere\nBaydhaba\nBerdaale\nDoolow\nKismaayo\nMogDaynile\nMogKahda\nXudur\n1%\n2%\n1%\n3%\n3%\n1%\n1%\n8%\n2%\n2%\n1%\n4%\n4%\n1%\n2%\n8%\nTrust in justice systems\nBaardheere\nBaydhaba\nBerdaale\nDoolow\nKismaayo\nMogDaynile\nMogKahda\nXudur\n68%\n77%\n78%\n83%\n93%\n84%\n84%\n86%\n67%\n75%\n97%\n76%\n85%\n75%\n82%\n87%\nTrust in local authorities\nBaardheere\nBaydhaba\nBerdaale\nDoolow\nKismaayo\nMogDaynile\nMogKahda\nXudur\n75%\n77%\n96%\n89%\n87%\n83%\n83%\n88%\n75%\n75%\n95%\n79%\n77%\n75%\n80%\n88%\nTrust in security services\nBaardheere\nBaydhaba\nBerdaale\nDoolow\nKismaayo\nMogDaynile\nMogKahda\nXudur\n72%\n84%\n94%\n88%\n93%\n89%\n89%\n91%\n74%\n81%\n97%\n84%\n86%\n81%\n86%\n88%\nGovernment Leadership\nSolutions Pathways Action Plan\n",
+    "datasets": [],
+    "document": {
+      "source": "/sample_pdfs/Somalia_Baseline.pdf",
+      "pages": [
+        5
+      ]
+    }
+  },
+  {
+    "input_text": "Power BI Desktop\nAccess to Basic \nServices\nGuaranteeing \nthat \nIDPs \nhave \nsustained \naccess \nto \nessential \nservices \nsuch \nas \nhealthcare, \neducation, water, and sanitation is \nvital to preventing dependence on \nhumanitarianassistance. \nSimultaneously, \nensuring \nthe \ninternally displaced are included in \ndeveloping \nnational \nsystems, \nincluding protection programmes \nand services and development \nstrategies, is vital for improving the \nliving conditions and well-being of \ncommunities.\nPathway #2:\n",
+    "datasets": [],
+    "document": {
+      "source": "/sample_pdfs/Somalia_Baseline.pdf",
+      "pages": [
+        6
+      ]
+    }
+  },
+  {
+    "input_text": "Power BI Desktop\nAccess to power source (public or private)\nBaardheere\nBaydhaba\nBerdaale\nDoolow\nKismaayo\nMogDaynile\nMogKahda\nXudur\n83%\n70%\n58%\n80%\n78%\n88%\n88%\n40%\n64%\n49%\n19%\n40%\n41%\n50%\n57%\n5%\nSafe access to latrines\nBaardheere\nBaydhaba\nBerdaale\nDoolow\nKismaayo\nMogDaynile\nMogKahda\nXudur\n89%\n89%\n93%\n85%\n94%\n91%\n91%\n78%\n52%\n74%\n75%\n82%\n71%\n90%\n66%\n81%\nSafe access to water source\nBaardheere\nBaydhaba\nBerdaale\nDoolow\nKismaayo\nMogDaynile\nMogKahda\nXudur\n91%\n88%\n90%\n89%\n89%\n96%\n96%\n92%\n78%\n86%\n88%\n91%\n70%\n94%\n83%\n83%\nAccess to sufficient water (dry season)\nBaardheere\nBaydhaba\nBerdaale\nDoolow\nKismaayo\nMogDaynile\nMogKahda\nXudur\n53%\n49%\n39%\n68%\n73%\n84%\n84%\n35%\n35%\n33%\n28%\n65%\n38%\n73%\n51%\n36%\nAccess to sufficient water (rainy season)\nBaardheere\nBaydhaba\nBerdaale\nDoolow\nKismaayo\nMogDaynile\nMogKahda\nXudur\n68%\n92%\n69%\n83%\n81%\n93%\n93%\n90%\n82%\n74%\n71%\n75%\n60%\n84%\n79%\n84%\nAccess to Basic Services\nTravel time to water source (less than 30 min)\nBaardheere\nBaydhaba\nBerdaale\nDoolow\nKismaayo\nMogDaynile\nMogKahda\nXudur\n70%\n75%\n65%\n91%\n87%\n91%\n91%\n78%\n44%\n70%\n48%\n91%\n82%\n86%\n81%\n78%\nAccess to Drinking Water\nAccess to Latrine \nAccess to Electricity \nAccess to Food \nSafe access to food sources\nBaardheere\nBaydhaba\nBerdaale\nDoolow\nKismaayo\nMogDaynile\nMogKahda\nXudur\n87%\n87%\n84%\n85%\n94%\n94%\n94%\n88%\n75%\n87%\n92%\n72%\n79%\n94%\n77%\n91%\nTravel time to food markets (less than 30 min)\nBaardheere\nBaydhaba\nBerdaale\nDoolow\nKismaayo\nMogDaynile\nMogKahda\nXudur\n56%\n66%\n84%\n88%\n83%\n82%\n82%\n90%\n30%\n63%\n45%\n81%\n73%\n82%\n69%\n83%\nSolutions Pathways Action Plan\n",
+    "datasets": [
+      {
+        "dataset_name": {
+          "text": "Baardheere Baydhaba Berdaale Doolow Kismaayo MogDaynile MogKahda Xudur",
+          "confidence": 0.3558395802974701,
+          "start": 1183,
+          "end": 1253,
+          "clean_text": "Baardheere Baydhaba Berdaale Doolow Kismaayo MogDaynile MogKahda Xudur"
+        },
+        "dataset_tag": "vague",
+        "description": null,
+        "data_type": "other",
+        "acronym": {
+          "text": "",
+          "confidence": 0.5230647921562195,
+          "start": null,
+          "end": null,
+          "clean_text": ""
+        },
+        "author": null,
+        "producer": {
+          "text": "",
+          "confidence": 0.0,
+          "start": 1183,
+          "end": 1253,
+          "clean_text": ""
+        },
+        "geography": {
+          "text": "",
+          "confidence": 0.0,
+          "start": 1183,
+          "end": 1253,
+          "clean_text": ""
+        },
+        "publication_year": null,
+        "reference_year": {
+          "text": "",
+          "confidence": 0.0,
+          "start": 1183,
+          "end": 1253
+        },
+        "reference_population": null,
+        "is_used": "False",
+        "usage_context": "background",
+        "validations": []
+      }
+    ],
+    "document": {
+      "source": "/sample_pdfs/Somalia_Baseline.pdf",
+      "pages": [
+        7
+      ]
+    }
+  },
+  {
+    "input_text": "Power BI Desktop\nReceived required healthcare (last 3 months)\nBaardheere\nBaydhaba\nBerdaale\nDoolow\nKismaayo\nMogDaynile\nMogKahda\nXudur\n65%\n63%\n65%\n87%\n74%\n87%\n87%\n87%\n43%\n72%\n44%\n74%\n58%\n61%\n62%\n86%\nSchool attendance rate (boys)\nBaardheere\nBaydhaba\nBerdaale\nDoolow\nKismaayo\nMogDaynile\nMogKahda\nXudur\n45%\n40%\n36%\n47%\n46%\n51%\n51%\n51%\n31%\n38%\n50%\n44%\n35%\n28%\n23%\n45%\nSchool attendance rate (girls)\nBaardheere\nBaydhaba\nBerdaale\nDoolow\nKismaayo\nMogDaynile\nMogKahda\nXudur\n45%\n37%\n40%\n49%\n42%\n45%\n45%\n52%\n26%\n36%\n50%\n38%\n37%\n23%\n19%\n39%\nTrust in education system\nBaardheere\nBaydhaba\nBerdaale\nDoolow\nKismaayo\nMogDaynile\nMogKahda\nXudur\n86%\n86%\n72%\n95%\n96%\n94%\n94%\n92%\n74%\n88%\n98%\n89%\n86%\n85%\n83%\n89%\nTrust in health system\nBaardheere\nBaydhaba\nBerdaale\nDoolow\nKismaayo\nMogDaynile\nMogKahda\nXudur\n93%\n89%\n87%\n97%\n93%\n96%\n96%\n96%\n86%\n90%\n98%\n92%\n90%\n92%\n90%\n95%\nAccess to Basic Services\nTravel time to health facilities (less than 30 min)\nBaardheere\nBaydhaba\nBerdaale\nDoolow\nKismaayo\nMogDaynile\nMogKahda\nXudur\n57%\n56%\n85%\n85%\n75%\n51%\n51%\n88%\n28%\n55%\n41%\n77%\n67%\n60%\n51%\n82%\nAccess to health facilities\nBaardheere\nBaydhaba\nBerdaale\nDoolow\nKismaayo\nMogDaynile\nMogKahda\nXudur\n96%\n92%\n100%\n100%\n98%\n100%\n100%\n100%\n96%\n98%\n100%\n100%\n98%\n99%\n99%\n100%\nTravel time to school (less than 30 min)\nBaardheere\nBaydhaba\nBerdaale\nDoolow\nKismaayo\nMogDaynile\nMogKahda\nXudur\n54%\n70%\n84%\n82%\n82%\n78%\n78%\n89%\n21%\n64%\n48%\n82%\n87%\n89%\n72%\n79%\nAccess to Health\nAccess to Education\nSolutions Pathways Action Plan\n",
+    "datasets": [
+      {
+        "dataset_name": {
+          "text": "Power BI Desktop Received required healthcare",
+          "confidence": 0.7799632549285889,
+          "start": 0,
+          "end": 45,
+          "clean_text": "Power BI Desktop Received required healthcare"
+        },
+        "dataset_tag": "named",
+        "description": null,
+        "data_type": "administrative",
+        "acronym": {
+          "text": "",
+          "confidence": 0.8571798205375671,
+          "start": null,
+          "end": null,
+          "clean_text": ""
+        },
+        "author": null,
+        "producer": {
+          "text": "Trust in education system",
+          "confidence": 0.7229317426681519,
+          "start": 528,
+          "end": 553,
+          "clean_text": "Trust in education system"
+        },
+        "geography": {
+          "text": "",
+          "confidence": 0.0,
+          "start": 0,
+          "end": 45,
+          "clean_text": ""
+        },
+        "publication_year": null,
+        "reference_year": {
+          "text": "last 3 months",
+          "confidence": 0.4956997036933899,
+          "start": 47,
+          "end": 60
+        },
+        "reference_population": null,
+        "is_used": "True",
+        "usage_context": "primary",
+        "validations": []
+      },
+      {
+        "dataset_name": {
+          "text": "Trust in health system",
+          "confidence": 0.4111464023590088,
+          "start": 689,
+          "end": 711,
+          "clean_text": "Trust in health system"
+        },
+        "dataset_tag": "named",
+        "description": null,
+        "data_type": "administrative",
+        "acronym": {
+          "text": "",
+          "confidence": 0.4652654826641083,
+          "start": null,
+          "end": null,
+          "clean_text": ""
+        },
+        "author": null,
+        "producer": {
+          "text": "",
+          "confidence": 0.0,
+          "start": 689,
+          "end": 711,
+          "clean_text": ""
+        },
+        "geography": {
+          "text": "",
+          "confidence": 0.0,
+          "start": 689,
+          "end": 711,
+          "clean_text": ""
+        },
+        "publication_year": null,
+        "reference_year": {
+          "text": "",
+          "confidence": 0.0,
+          "start": 689,
+          "end": 711
+        },
+        "reference_population": null,
+        "is_used": "True",
+        "usage_context": "primary",
+        "validations": []
+      }
+    ],
+    "document": {
+      "source": "/sample_pdfs/Somalia_Baseline.pdf",
+      "pages": [
+        8
+      ]
+    }
+  },
+  {
+    "input_text": "Power BI Desktop\nEmployments\n&\nLivelihood\nOpportunities\nFacilitating \naccess \nto \njob \nopportunities \nand \nsustainable \nlivelihoods is critical to ensuring \nresilient communities. Providing \nIDPs with the means to support \nthemselves \neconomically \ncan \nsignificantly enhance their ability to \nrebuild their lives and reduce \ndependency on humanitarian aid.\nPathway #3:\n",
+    "datasets": [],
+    "document": {
+      "source": "/sample_pdfs/Somalia_Baseline.pdf",
+      "pages": [
+        9
+      ]
+    }
+  },
+  {
+    "input_text": "Power BI Desktop\nAble to meet all/some basic needs independently\nBaardheere\nBaydhaba\nBerdaale\nDoolow\nKismaayo\nMogDaynile\nMogKahda\nXudur\n64%\n81%\n85%\n89%\n76%\n89%\n89%\n80%\n60%\n76%\n35%\n64%\n49%\n67%\n72%\n46%\nAccess to financial institutions\nBaardheere\nBaydhaba\nBerdaale\nDoolow\nKismaayo\nMogDaynile\nMogKahda\nXudur\n12%\n22%\n8%\n9%\n13%\n21%\n21%\n15%\n16%\n8%\n4%\n1%\n0%\n1%\n1%\n7%\nAccess to credit options\nBaardheere\nBaydhaba\nBerdaale\nDoolow\nKismaayo\nMogDaynile\nMogKahda\nXudur\n10%\n7%\n1%\n3%\n7%\n5%\n5%\n1%\n3%\n2%\n1%\n2%\n0%\n1%\n2%\n0%\nHouseholds with no debt (except investments)\nBaardheere\nBaydhaba\nBerdaale\nDoolow\nKismaayo\nMogDaynile\nMogKahda\nXudur\n1%\n17%\n8%\n12%\n19%\n12%\n12%\n4%\n1%\n7%\n5%\n12%\n12%\n8%\n8%\n1%\nEmployment and Livelihood\nSolutions Pathways Action Plan\nDifferent types of climate adaptation measures employed\nCity\nBaardheere\nBaydhaba\nBerdaale\nDoolow\nKismaayo\nMogDaynile\nMogKahda\nXudur\nIncome\nHC\nIDP\nHC\nIDP\nHC\nIDP\nHC\nIDP\nHC\nIDP\nHC\nIDP\nHC\nIDP\nHC\nIDP\nHousehold’s crops\n10%\n12%\n10%\n8%\n8%\n16%\n4%\n4%\n2%\n0%\n1%\n1%\n1%\n3%\n11%\n8%\nOwn business or commerce\n4%\n1%\n5%\n3%\n7%\n0%\n10%\n3%\n6%\n1%\n11%\n1%\n11%\n1%\n7%\n0%\nSalaried work\n3%\n2%\n9%\n3%\n3%\n1%\n6%\n0%\n13%\n1%\n17%\n2%\n17%\n3%\n7%\n2%\nSelling animal products\n9%\n7%\n10%\n5%\n2%\n10%\n3%\n3%\n1%\n1%\n3%\n2%\n3%\n3%\n3%\n1%\nSelling animals\n18%\n7%\n13%\n10%\n6%\n7%\n4%\n2%\n3%\n1%\n3%\n3%\n3%\n2%\n4%\n4%\n",
+    "datasets": [],
+    "document": {
+      "source": "/sample_pdfs/Somalia_Baseline.pdf",
+      "pages": [
+        10
+      ]
+    }
+  },
+  {
+    "input_text": "Power BI Desktop\nLegal Documentation, \nHousing, Land, & \nProperty (HLP), and \nAccess to Justice\nProviding necessary documentation and ensuring access to justice, \nhousing, land, and property rights are essential for IDPs. Legal \ndocumentation is often necessary to access basic and financial services \nand secure employment, whilst tenure security and housing and land \nrights are critical for achieving permanent housing solutions to \ndisplacement.\nPathway #4:\n",
+    "datasets": [],
+    "document": {
+      "source": "/sample_pdfs/Somalia_Baseline.pdf",
+      "pages": [
+        11
+      ]
+    }
+  },
+  {
+    "input_text": "Power BI Desktop\nResiding in permanent or transitional shelters\nBaardheere\nBaydhaba\nBerdaale\nDoolow\nKismaayo\nMogDaynile\nMogKahda\nXudur\n89%\n85%\n56%\n83%\n94%\n89%\n89%\n95%\n10%\n42%\n2%\n17%\n39%\n25%\n17%\n4%\nPossession of legal identity documentation\nBaardheere\nBaydhaba\nBerdaale\nDoolow\nKismaayo\nMogDaynile\nMogKahda\nXudur\n49%\n27%\n11%\n48%\n36%\n33%\n33%\n16%\n14%\n15%\n1%\n15%\n19%\n2%\n6%\n3%\nAccess to effective justice systems\nBaardheere\nBaydhaba\nBerdaale\nDoolow\nKismaayo\nMogDaynile\nMogKahda\nXudur\n91%\n91%\n93%\n96%\n96%\n94%\n94%\n95%\n90%\n98%\n95%\n98%\n98%\n84%\n92%\n94%\nLegal Documentation, and HLP\nPossession of land documentation\nBaardheere\nBaydhaba\nBerdaale\nDoolow\nKismaayo\nMogDaynile\nMogKahda\nXudur\n34%\n58%\n36%\n67%\n50%\n25%\n25%\n56%\n9%\n50%\n9%\n13%\n13%\n6%\n11%\n11%\nAccess to Documentation\nAccess to Shelter\nRisk of Eviction\nNot facing eviction risk\nBaardheere\nBaydhaba\nBerdaale\nDoolow\nKismaayo\nMogDaynile\nMogKahda\nXudur\n49%\n66%\n39%\n74%\n71%\n51%\n51%\n66%\n32%\n56%\n52%\n81%\n63%\n27%\n47%\n50%\nNo involvement in land disputes\nBaardheere\nBaydhaba\nBerdaale\nDoolow\nKismaayo\nMogDaynile\nMogKahda\nXudur\n89%\n83%\n87%\n86%\n85%\n93%\n93%\n84%\n90%\n80%\n94%\n93%\n87%\n97%\n73%\n91%\nSolutions Pathways Action Plan\nAccess to legal services\nBaardheere\nBaydhaba\nBerdaale\nDoolow\nKismaayo\nMogDaynile\nMogKahda\nXudur\n56%\n65%\n36%\n74%\n73%\n60%\n60%\n80%\n49%\n68%\n48%\n62%\n63%\n54%\n32%\n85%\n",
+    "datasets": [],
+    "document": {
+      "source": "/sample_pdfs/Somalia_Baseline.pdf",
+      "pages": [
+        12
+      ]
+    }
+  },
+  {
+    "input_text": "Power BI Desktop\nClimate Change \n& Resilience\nEnhancing the climate adaption \nand resilience of IDPs and host \ncommunities \nis \na \npriority. \nStrengthening their ability to cope \nwith and recover from current and \nfuture shocks, such as conflicts or \nclimate \nchange \ndisasters, \nis \nessential \nfor \nsustainable \ndevelopment and reducing the risk \nof further displacement.\nPathway #5:\n",
+    "datasets": [],
+    "document": {
+      "source": "/sample_pdfs/Somalia_Baseline.pdf",
+      "pages": [
+        13
+      ]
+    }
+  },
+  {
+    "input_text": "Power BI Desktop\nEmployment of any type of climate adaptation measures\nBaardheere\nBaydhaba\nBerdaale\nDoolow\nKismaayo\nMogDaynile\nMogKahda\nXudur\n24%\n48%\n17%\n14%\n39%\n26%\n26%\n12%\n31%\n48%\n39%\n25%\n33%\n19%\n15%\n18%\nClimate Change & Resilience\nSolutions Pathways Action Plan\nThis analysis highlights significant differences in climate adaptation measures between Host Communities (HCs) and \nInternally Displaced Persons (IDPs) within the same cities in Somalia. IDPs consistently report higher levels of adaptation \nand adopt a broader range of strategies, largely driven by their heightened vulnerability and displacement experiences. In \ncontrast, HCs leverage their stable housing and controlled environments to focus on structural measures such as improving \nbuildings and drainage systems.\nIDPs often rely on adaptive strategies like migration and temporary protective actions, reflecting their need for immediate \nand flexible responses to environmental challenges. However, gaps in adaptation efforts, particularly in cities like Xudur, \nDoolow, \nand \nMogadishu, \nhighlight \nareas \nwhere \nboth \ngroups \nrequire \ntargeted \ninterventions.\nThis contrast in adaptation approaches underscores the need for tailored support that considers the unique circumstances \nand strengths of each group, fostering resilience to climate change across Somalia.\nDifferent types of climate adaptation measures employed\nCity\nBaardheere\nBaydhaba\nBerdaale\nDoolow\nKismaayo\nMogDaynile\nMogKahda\nXudur\nBreakdown\n \nHC\nIDP\nHC\nIDP\nHC\nIDP\nHC\nIDP\nHC\nIDP\nHC\nIDP\nHC\nIDP\nHC\nIDP\nClear drainage\n47%\n23%\n17%\n21%\n25%\n7%\n43%\n47%\n37%\n34%\n24%\n7%\n24%\n34%\n13%\n3%\nConstruct physical barriers\n0%\n0%\n16%\n19%\n11%\n4%\n24%\n8%\n18%\n3%\n5%\n8%\n5%\n5%\n39%\n17%\nConstruct stronger roofs\n1%\n4%\n18%\n21%\n14%\n8%\n41%\n15%\n26%\n15%\n12%\n18%\n12%\n2%\n55%\n28%\nDig trenches around the house\n15%\n11%\n22%\n27%\n21%\n9%\n10%\n11%\n47%\n57%\n38%\n47%\n38%\n6%\n18%\n22%\nDiversify economic activity\n0%\n1%\n7%\n6%\n0%\n0%\n10%\n0%\n3%\n2%\n2%\n2%\n2%\n0%\n3%\n1%\nIrrigation of dry land\n1%\n1%\n4%\n9%\n25%\n19%\n0%\n2%\n0%\n0%\n0%\n0%\n0%\n0%\n5%\n0%\nMigration to another location\n37%\n27%\n41%\n42%\n57%\n55%\n2%\n2%\n22%\n17%\n25%\n23%\n25%\n54%\n39%\n70%\nReduce consumption of food\n31%\n37%\n8%\n18%\n7%\n20%\n35%\n70%\n5%\n2%\n32%\n27%\n32%\n8%\n8%\n1%\nSending family members elsewhere\n1%\n4%\n7%\n17%\n36%\n37%\n12%\n4%\n1%\n0%\n6%\n15%\n6%\n1%\n26%\n21%\nUse stronger building materials\n1%\n2%\n15%\n15%\n4%\n8%\n8%\n3%\n1%\n1%\n2%\n7%\n2%\n0%\n39%\n7%\n",
+    "datasets": [],
+    "document": {
+      "source": "/sample_pdfs/Somalia_Baseline.pdf",
+      "pages": [
+        14
+      ]
+    }
+  },
+  {
+    "input_text": "Power BI Desktop\nSolutions Pathways Action Plan\nProtection and Conflict Sensitivity\n-\nBaardheere: A high occurence of violence and sexual and gender-based \nviolence (SGBV) perpretrated by non-state actors as well as private individuals, \nhighlighting a complicated conflict landscape. There is a need for strengthened \nprotection services and provision of basic services.\n-\nBaydhaba: Findings suggest high rates of sexual and gender based violence \n(SGBV) took place, and were mostly perpretrated by non-state actors and private \nindividuals. There is a need for enhanced protection from violence, and provision \nof basic services such as food and shelter.\n-\nBerdaale: Notably high incidents of violence and SGBV were recorded, \nmostly by non-state actors and private individuals, highlighting localized conflict \ndynamics. There is a need for strengthened protection for vulnerable groups.\n-\nDoolow: Escalating conflict was highlighted, with a high occurrence of \nviolence and SGBV perpetrated by non-state actors, alongside private individuals. \nThere is a need for strengthened security services.\n-\nKismayo: A high level of violence and killings were reported, perpetrated \nby private actors, which suggests the prevalence of opportunistic violence with \nunaffiliated private groups exploiting insecurity. There is a need for strengthened \nlaw enforcement capacity, and community-based protection mechanisms.\n-\nBanadir (including Daynile and Kahda): Very high displacements were \nrecorded due to violence, SGBV, and killings with a combination of actors \ninvolved - predominantly unaffiliated private individuals and non-state actors. \nThere is a need for strengthened law enforcement capacity and a prioritization of \nbasic services as well as livelihoods support to meet the needs of displaced \ncommunities.\n-\nXudur: High rates of SGBV, exploitation and child protection violences \nwere recorded, with penetrators identified as unaffiliated private actors, \nsuggesting a limited state presence. There is a need to strengthen community \nmechanisms which are commonly relied upon by affected populations. There is \nalso a need to improve access to food and shelter, and address exploitation \nthrough targeted protection measures.\nIn 2024, UNHCR conducted several assessment and analysis activities focused on \nprotection and conflict sensitivity analysis. Key findings from these activities in the eight \nlocations referred to in this report are detailed below. Note that analysis from the whole \nyear is summarised.\n",
+    "datasets": [],
+    "document": {
+      "source": "/sample_pdfs/Somalia_Baseline.pdf",
+      "pages": [
+        15
+      ]
+    }
+  },
+  {
+    "input_text": "Power BI Desktop\n",
+    "datasets": [],
+    "document": {
+      "source": "/sample_pdfs/Somalia_Baseline.pdf",
+      "pages": [
+        16
+      ]
+    }
+  }
+]

annotation_data/sample_extractions/doc_12/raw/doc_12_direct_judged.jsonl

@@ -0,0 +1,480 @@
+[
+  {
+    "input_text": "[image]\nSites\n•  IOM Global\n•  GDI\nENESFR\nUser account menu\nLog in Contact\n•  About Toggle submenu\n○  About DTM\n○  DTM Data Kit\n○  Updates\n○  Methodological Framework\n○  DTM Standards\n○  Infosheets\n○  Trainings\n○  Vacancies\n○  Global Data Institute\n•  Countries Toggle submenu\n○  Asia and the Pacific Toggle submenu\n■  Afghanistan\n■  Bangladesh\n■  Cambodia\n■  Federated States of Micronesia\n■  Fiji\n■  Indonesia\n■  Iran (Islamic Republic of)\n■  Laos\n■  Marshall Islands\n",
+    "datasets": [
+      {
+        "dataset_name": {
+          "text": "IOM Global",
+          "confidence": 0.5996029376983643,
+          "start": 16,
+          "end": 26,
+          "clean_text": "IOM Global"
+        },
+        "dataset_tag": "named",
+        "description": null,
+        "data_type": "other",
+        "acronym": {
+          "text": "",
+          "confidence": 0.9287746548652649,
+          "start": null,
+          "end": null,
+          "clean_text": ""
+        },
+        "author": null,
+        "producer": {
+          "text": "IOM",
+          "confidence": 0.7735210061073303,
+          "start": 16,
+          "end": 19,
+          "clean_text": "IOM"
+        },
+        "geography": {
+          "text": "Afghanistan",
+          "confidence": 0.8836480379104614,
+          "start": 304,
+          "end": 315,
+          "clean_text": "Afghanistan"
+        },
+        "publication_year": null,
+        "reference_year": {
+          "text": "",
+          "confidence": 0.0,
+          "start": 16,
+          "end": 26
+        },
+        "reference_population": null,
+        "is_used": "True",
+        "usage_context": "primary",
+        "validations": []
+      },
+      {
+        "dataset_name": {
+          "text": "DTM Data Kit",
+          "confidence": 0.8222224116325378,
+          "start": 110,
+          "end": 122,
+          "clean_text": "DTM Data Kit"
+        },
+        "dataset_tag": "named",
+        "description": null,
+        "data_type": "other",
+        "acronym": {
+          "text": "",
+          "confidence": 0.0,
+          "start": 110,
+          "end": 122,
+          "clean_text": ""
+        },
+        "author": null,
+        "producer": {
+          "text": "Global Data Institute",
+          "confidence": 0.559898853302002,
+          "start": 215,
+          "end": 236,
+          "clean_text": "Global Data Institute"
+        },
+        "geography": {
+          "text": "Cambodia",
+          "confidence": 0.5714520215988159,
+          "start": 331,
+          "end": 339,
+          "clean_text": "Cambodia"
+        },
+        "publication_year": null,
+        "reference_year": {
+          "text": "",
+          "confidence": 0.0,
+          "start": 110,
+          "end": 122
+        },
+        "reference_population": null,
+        "is_used": "True",
+        "usage_context": "primary",
+        "validations": []
+      }
+    ],
+    "document": {
+      "source": "/sample_pdfs/Somalia_NSP.pdf",
+      "pages": [
+        0
+      ]
+    }
+  },
+  {
+    "input_text": "■  Mongolia\n■  Myanmar\n■  Nepal\n■  Pakistan\n■  Papua New Guinea\n■  Philippines\n■  Solomon Islands\n■  Sri Lanka\n■  Thailand\n■  Tonga\n■  Vanuatu\n○  Latin America and the Caribbean Toggle submenu\n■  Antigua and Barbuda\n■  Argentina\n■  Aruba\n■  Bahamas\n■  Bolivia\n■  Brazil\n■  Chile\n■  Colombia\n■  Costa Rica\n■  Curaçao\n■  Dominica\n■  Dominican Republic\n■  Ecuador\n■  El Salvador\n■  Grenada\n■  Guatemala\n■  Guyana\n■  Haiti\n■  Honduras\n■  Jamaica\n■  Mexico\n■  Nicaragua\n■  Panama\n■  Paraguay\n",
+    "datasets": [],
+    "document": {
+      "source": "/sample_pdfs/Somalia_NSP.pdf",
+      "pages": [
+        1
+      ]
+    }
+  },
+  {
+    "input_text": "■  Peru\n■  Saint Vincent and the Grenadines\n■  Sint Maarten\n■  Surinam\n■  Trinidad & Tobago\n■  Uruguay\n■  Venezuela\n○  East, Horn and Southern Africa Toggle submenu\n■  Burundi\n■  Democratic Republic of the Congo\n■  Djibouti\n■  Ethiopia\n■  Kenya\n■  Lesotho\n■  Madagascar\n■  Malawi\n■  Mozambique\n■  Namibia\n■  Somalia\n■  South Africa\n■  South Sudan\n■  Uganda\n■  United Republic of Tanzania\n■  Zambia\n■  Zimbabwe\n○  Europe and Central Asia Toggle submenu\n■  Albania\n■  Armenia\n■  Austria\n■  Belarus\n■  Belgium\n■  Bosnia and Herzegovina\n■  Bulgaria\n■  Croatia\n■  Cyprus\n■  Czechia\n",
+    "datasets": [
+      {
+        "dataset_name": {
+          "text": "Saint Vincent and the Grenadines",
+          "confidence": 0.7407812476158142,
+          "start": 9,
+          "end": 41,
+          "clean_text": "Saint Vincent and the Grenadines"
+        },
+        "dataset_tag": "named",
+        "description": null,
+        "data_type": "other",
+        "acronym": {
+          "text": "",
+          "confidence": 0.7671184539794922,
+          "start": null,
+          "end": null,
+          "clean_text": ""
+        },
+        "author": null,
+        "producer": {
+          "text": "",
+          "confidence": 0.0,
+          "start": 9,
+          "end": 41,
+          "clean_text": ""
+        },
+        "geography": {
+          "text": "Peru",
+          "confidence": 0.8073104619979858,
+          "start": 2,
+          "end": 6,
+          "clean_text": "Peru"
+        },
+        "publication_year": null,
+        "reference_year": {
+          "text": "",
+          "confidence": 0.0,
+          "start": 9,
+          "end": 41
+        },
+        "reference_population": null,
+        "is_used": "True",
+        "usage_context": "primary",
+        "validations": []
+      },
+      {
+        "dataset_name": {
+          "text": "East, Horn and Southern Africa Toggle submenu",
+          "confidence": 0.5781986117362976,
+          "start": 111,
+          "end": 156,
+          "clean_text": "East, Horn and Southern Africa Toggle submenu"
+        },
+        "dataset_tag": "named",
+        "description": null,
+        "data_type": "other",
+        "acronym": {
+          "text": "",
+          "confidence": 0.0,
+          "start": 111,
+          "end": 156,
+          "clean_text": ""
+        },
+        "author": null,
+        "producer": {
+          "text": "",
+          "confidence": 0.0,
+          "start": 111,
+          "end": 156,
+          "clean_text": ""
+        },
+        "geography": {
+          "text": "Burundi",
+          "confidence": 0.510351836681366,
+          "start": 159,
+          "end": 166,
+          "clean_text": "Burundi"
+        },
+        "publication_year": null,
+        "reference_year": {
+          "text": "",
+          "confidence": 0.0,
+          "start": 111,
+          "end": 156
+        },
+        "reference_population": null,
+        "is_used": "True",
+        "usage_context": "primary",
+        "validations": []
+      }
+    ],
+    "document": {
+      "source": "/sample_pdfs/Somalia_NSP.pdf",
+      "pages": [
+        2
+      ]
+    }
+  },
+  {
+    "input_text": "■  Estonia\n■  France\n■  Germany\n■  Greece\n■  Hungary\n■  Italy\n■  Kazakhstan\n■  Kosovo 1244\n■  Kyrgyzstan\n■  Latvia\n■  Lithuania\n■  Malta\n■  Montenegro\n■  Netherlands\n■  North Macedonia\n■  Poland\n■  Republic of Moldova\n■  Republic of Türkiye\n■  Romania\n■  Serbia\n■  Slovakia\n■  Slovenia\n■  Spain\n■  Sweden\n■  Tajikistan\n■  Ukraine\n■  Uzbekistan\n○  Middle East and North Africa Toggle submenu\n■  Iraq\n■  Lebanon\n■  Libya\n■  Sudan\n■  Syrian Arab Republic\n■  Tunisia\n■  Yemen\n○  West and Central Africa Toggle submenu\n",
+    "datasets": [],
+    "document": {
+      "source": "/sample_pdfs/Somalia_NSP.pdf",
+      "pages": [
+        3
+      ]
+    }
+  },
+  {
+    "input_text": "■  Benin\n■  Burkina Faso\n■  Cameroon\n■  Central African Republic\n■  Chad\n■  Côte d’Ivoire\n■  Ghana\n■  Guinea\n■  Guinea-Bissau\n■  Liberia\n■  Mali\n■  Mauritania\n■  Niger\n■  Nigeria\n■  Senegal\n■  Sierra Leone\n■  The Gambia\n■  Togo\n•  Data and Analysis Toggle submenu\n○  Reports\n○  Datasets\n○  API\n○  Maps\n○  Online Interactive Resources\n•  Toolkit\n•  Special Focus Toggle submenu\n○  Migrant Response Plan\n○  Climate and Mobility\n○  Data for solutions Toggle submenu\n■  PROGRESS 2025\n■  PROGRESS 2024\n■  PROGRESS 2023\n■  Solutions and Mobility Index\n○  Youth Engagement\n",
+    "datasets": [
+      {
+        "dataset_name": {
+          "text": "Data and Analysis Toggle submenu",
+          "confidence": 0.7612807750701904,
+          "start": 212,
+          "end": 244,
+          "clean_text": "Data and Analysis Toggle submenu"
+        },
+        "dataset_tag": "named",
+        "description": null,
+        "data_type": "other",
+        "acronym": {
+          "text": "",
+          "confidence": 0.0,
+          "start": 212,
+          "end": 244,
+          "clean_text": ""
+        },
+        "author": null,
+        "producer": {
+          "text": "API",
+          "confidence": 0.6447415351867676,
+          "start": 268,
+          "end": 271,
+          "clean_text": "API"
+        },
+        "geography": {
+          "text": "Burkina Faso",
+          "confidence": 0.9339057803153992,
+          "start": 10,
+          "end": 22,
+          "clean_text": "Burkina Faso"
+        },
+        "publication_year": null,
+        "reference_year": {
+          "text": "",
+          "confidence": 0.0,
+          "start": 212,
+          "end": 244
+        },
+        "reference_population": null,
+        "is_used": "True",
+        "usage_context": "primary",
+        "validations": []
+      },
+      {
+        "dataset_name": {
+          "text": "API",
+          "confidence": 0.500595211982727,
+          "start": 268,
+          "end": 271,
+          "clean_text": "API"
+        },
+        "dataset_tag": "vague",
+        "description": null,
+        "data_type": "other",
+        "acronym": {
+          "text": "",
+          "confidence": 0.9450010657310486,
+          "start": null,
+          "end": null,
+          "clean_text": ""
+        },
+        "author": null,
+        "producer": {
+          "text": "",
+          "confidence": 0.0,
+          "start": 268,
+          "end": 271,
+          "clean_text": ""
+        },
+        "geography": {
+          "text": "",
+          "confidence": 0.0,
+          "start": 268,
+          "end": 271,
+          "clean_text": ""
+        },
+        "publication_year": null,
+        "reference_year": {
+          "text": "",
+          "confidence": 0.0,
+          "start": 268,
+          "end": 271
+        },
+        "reference_population": null,
+        "is_used": "False",
+        "usage_context": "background",
+        "validations": []
+      },
+      {
+        "dataset_name": {
+          "text": "Maps",
+          "confidence": 0.48470646142959595,
+          "start": 274,
+          "end": 278,
+          "clean_text": "Maps"
+        },
+        "dataset_tag": "vague",
+        "description": null,
+        "data_type": "other",
+        "acronym": {
+          "text": "",
+          "confidence": 0.0,
+          "start": 274,
+          "end": 278,
+          "clean_text": ""
+        },
+        "author": null,
+        "producer": {
+          "text": "",
+          "confidence": 0.0,
+          "start": 274,
+          "end": 278,
+          "clean_text": ""
+        },
+        "geography": {
+          "text": "",
+          "confidence": 0.0,
+          "start": 274,
+          "end": 278,
+          "clean_text": ""
+        },
+        "publication_year": null,
+        "reference_year": {
+          "text": "",
+          "confidence": 0.0,
+          "start": 274,
+          "end": 278
+        },
+        "reference_population": null,
+        "is_used": "False",
+        "usage_context": "background",
+        "validations": []
+      },
+      {
+        "dataset_name": {
+          "text": "Special Focus Toggle submenu",
+          "confidence": 0.6337509751319885,
+          "start": 322,
+          "end": 350,
+          "clean_text": "Special Focus Toggle submenu"
+        },
+        "dataset_tag": "vague",
+        "description": null,
+        "data_type": "other",
+        "acronym": {
+          "text": "",
+          "confidence": 0.0,
+          "start": 322,
+          "end": 350,
+          "clean_text": ""
+        },
+        "author": null,
+        "producer": {
+          "text": "",
+          "confidence": 0.0,
+          "start": 322,
+          "end": 350,
+          "clean_text": ""
+        },
+        "geography": {
+          "text": "",
+          "confidence": 0.0,
+          "start": 322,
+          "end": 350,
+          "clean_text": ""
+        },
+        "publication_year": null,
+        "reference_year": {
+          "text": "",
+          "confidence": 0.0,
+          "start": 322,
+          "end": 350
+        },
+        "reference_population": null,
+        "is_used": "False",
+        "usage_context": "background",
+        "validations": []
+      },
+      {
+        "dataset_name": {
+          "text": "Data for solutions Toggle submenu",
+          "confidence": 0.8023719191551208,
+          "start": 400,
+          "end": 433,
+          "clean_text": "Data for solutions Toggle submenu"
+        },
+        "dataset_tag": "vague",
+        "description": null,
+        "data_type": "other",
+        "acronym": {
+          "text": "",
+          "confidence": 0.0,
+          "start": 400,
+          "end": 433,
+          "clean_text": ""
+        },
+        "author": null,
+        "producer": {
+          "text": "",
+          "confidence": 0.0,
+          "start": 400,
+          "end": 433,
+          "clean_text": ""
+        },
+        "geography": {
+          "text": "Chad",
+          "confidence": 0.7367566823959351,
+          "start": 63,
+          "end": 67,
+          "clean_text": "Chad"
+        },
+        "publication_year": null,
+        "reference_year": {
+          "text": "",
+          "confidence": 0.0,
+          "start": 400,
+          "end": 433
+        },
+        "reference_population": null,
+        "is_used": "False",
+        "usage_context": "background",
+        "validations": []
+      }
+    ],
+    "document": {
+      "source": "/sample_pdfs/Somalia_NSP.pdf",
+      "pages": [
+        4
+      ]
+    }
+  },
+  {
+    "input_text": "•  Crisis Responses Toggle submenu\n○  Escalation in the Middle East and Beyond - Mobility\nMonitoring\n○  Ukraine Response\n○  Sudan Crisis\n○  Migration Flow to Europe\n○  COVID-19 Mobility Impact Toggle submenu\n■  Methodology\n■  Reports\n■  Sitreps\n■  Maps\n○  Returns and displacement in Afghanistan\n○  Migration Flows from Venezuela to the Americas\n○  Lake Chad Basin Crisis\n•  DTM INSIGHTS\n1. Home\nPage not found\nShare\nThe requested page could not be found.\n[image]\n•  \n•  \n•  \n•  \n•  \n•  \nHeadquarters - DTM Global Support\n",
+    "datasets": [],
+    "document": {
+      "source": "/sample_pdfs/Somalia_NSP.pdf",
+      "pages": [
+        5
+      ]
+    }
+  },
+  {
+    "input_text": "Rte des Morillons 17,\n1218 Le Grand-Saconnex,\nSwitzerland\nEmail: dtmsupport@iom.int\nContact\n[image]\nSubscribe to our mailing list here\nor by scanning the QR code above\nRelated Sites\n•  International Organization for Migration\n•  Global Data Institute\n•  Responding to COVID-19\n•  International Organization for Migration - HDX\n•  Displacement Tracking Matrix\n•  Migrants Arrival to Europe\nUseful Links\n•  Reports\n•  Datasets\n•  Maps\n•  Methodological Frameworks\n•  DTM Standards\n•  Infosheets\n•  Trainings\n•  Vacancies\nCopyright © 2026 International Organization for Migration\n•  Cyber-security Issues\n•  Terms & Conditions\n",
+    "datasets": [],
+    "document": {
+      "source": "/sample_pdfs/Somalia_NSP.pdf",
+      "pages": [
+        6
+      ]
+    }
+  }
+]

annotation_data/sample_extractions/doc_2/raw/doc_2_direct_judged.jsonl

@@ -248,51 +248,6 @@
         "usage_context": "primary",
         "validations": []
       },
-      {
-        "dataset_name": {
-          "text": "Eurostat’s at-risk-of-poverty (AROP)",
-          "confidence": 0.6725184321403503,
-          "start": 3343,
-          "end": 3379,
-          "clean_text": "Eurostat’s at-risk-of-poverty (AROP)"
-        },
-        "dataset_tag": "named",
-        "description": null,
-        "data_type": "other",
-        "acronym": {
-          "text": "",
-          "confidence": 0.9905750751495361,
-          "start": null,
-          "end": null,
-          "clean_text": ""
-        },
-        "author": null,
-        "producer": {
-          "text": "Eurostat",
-          "confidence": 0.9854599833488464,
-          "start": 3546,
-          "end": 3554,
-          "clean_text": "Eurostat"
-        },
-        "geography": {
-          "text": "",
-          "confidence": 0.0,
-          "start": 3343,
-          "end": 3379,
-          "clean_text": ""
-        },
-        "publication_year": null,
-        "reference_year": {
-          "text": "",
-          "confidence": 0.0,
-          "start": 3343,
-          "end": 3379
-        },
-        "reference_population": null,
-        "is_used": "True",
-        "usage_context": "supporting",
-        "validations": []
-      },
       {
         "dataset_name": {
           "text": "MHPSS",
@@ -349,51 +304,6 @@
   {
     "input_text": "4\t\nHIGH EMPLOYMENT RATES, BUT LOW WAGES: A POVERTY ASSESSMENT OF UKRAINIAN REFUGEES IN NEIGHBORING COUNTRIES\nBackground\nOver three years have elapsed since the start of the full-scale invasion of Ukraine, an event which has led to \nthe largest displacement crisis in Europe since World War II. As of the end of 2024, 6.3 million refugees from \nUkraine were recorded across Europe, close to 2 million of whom are located in ten Regional Refugee \nResponse Plan countries: Bulgaria, Czechia, Estonia, Hungary, Latvia, Lithuania, Poland, Republic of Moldova, \nRomania, and Slovakia. This report aims to assess the livelihood situation of this population based on the 2024 \nround of data collected by the Socio-Economic Insights Survey (SEIS), which received responses from 8,723 \nhouseholds containing 19,803 individuals. Figures for 2023 are derived from a similar exercise7 conducted in \n2023.\nKey findings\nRefugee poverty rates remain high, albeit improved from 2023\nThe 2024 SEIS equivalized8 disposable income data indicates that just over one in five refugees (23%) residing \nin the region are living in poverty9. This figure is almost double that of host country nationals (12%), implying a \nlarge gap in economic vulnerability. Compared to 2023, poverty rates have decreased substantially (from \n36%10), suggesting an overall improvement in the economic well-being of Ukrainian refugees over time. \n7.\t\nThe MSNA, which ran in 7 countries: Bulgaria, Czech Republic, Hungary, Republic of Moldova, Poland, Romania, and Slovakia\n8.\t Equivalized as per Eurostat methodology. Essentially income per person, but with household members beyond the first one \nassigned weights less than 1.\n9.\t With the poverty line defined as 50% of the median national equivalized income \n10.\t This figure differs from the one reported in the previous year’s publication due to changes to methodology, which includes a \ndifferent treatment of households that report zero income \nREFUGEE VERSUS HOST POVERTY RATES BY COUNTRY\nBulgaria\nCzechia\nHungary\nMoldova\nPoland\nRomania\nSlovakia\nEstonia\nLatvia\nLithuania\nRegion\n36%\n22%\n32%\n14%\n20%\n18%\n38%\n27%\n51%\n41%\n23%\n12%\n5%\n9%\n32%\n9%\n17%\n10%\n15%\n15%\n13%\n12%\nUkrainian refugees (2024)\nHost country nationals (2023)\nNote: Poverty rates for all countries apart from the Republic of Moldova are based on a calculation that follows Eurostat’s at-risk-of-poverty (AROP) \nmethodology with the at-risk-of-poverty threshold set at 50% of the national median disposable income after social transfers. Refugee disposable \nincome has been computed based on survey data. For the Republic of Moldova, the poverty threshold was taken to be the 4Q23 absolute poverty line \nreported by the National Bureau of Statistics of Moldova. \nSource: Survey data, Eurostat, National Bureau of Statistics of Moldova, SAG estimates\n",
     "datasets": [
-      {
-        "dataset_name": {
-          "text": "POVERTY ASSESSMENT OF UKRAINIAN REFUGEES IN NEIGHBORING COUNTRIES",
-          "confidence": 0.3901032507419586,
-          "start": 42,
-          "end": 107,
-          "clean_text": "POVERTY ASSESSMENT OF UKRAINIAN REFUGEES IN NEIGHBORING COUNTRIES"
-        },
-        "dataset_tag": "named",
-        "description": null,
-        "data_type": "other",
-        "acronym": {
-          "text": "",
-          "confidence": 0.0,
-          "start": 42,
-          "end": 107,
-          "clean_text": ""
-        },
-        "author": null,
-        "producer": {
-          "text": "",
-          "confidence": 0.0,
-          "start": 42,
-          "end": 107,
-          "clean_text": ""
-        },
-        "geography": {
-          "text": "",
-          "confidence": 0.0,
-          "start": 42,
-          "end": 107,
-          "clean_text": ""
-        },
-        "publication_year": null,
-        "reference_year": {
-          "text": "",
-          "confidence": 0.0,
-          "start": 42,
-          "end": 107
-        },
-        "reference_population": null,
-        "is_used": "True",
-        "usage_context": "supporting",
-        "validations": []
-      },
       {
         "dataset_name": {
           "text": "Socio-Economic Insights Survey (SEIS)",
@@ -529,51 +439,6 @@
         "usage_context": "primary",
         "validations": []
       },
-      {
-        "dataset_name": {
-          "text": "Ukrainian refugees (2024)",
-          "confidence": 0.3187820613384247,
-          "start": 2160,
-          "end": 2185,
-          "clean_text": "Ukrainian refugees (2024)"
-        },
-        "dataset_tag": "vague",
-        "description": null,
-        "data_type": "other",
-        "acronym": {
-          "text": "",
-          "confidence": 0.0,
-          "start": 2160,
-          "end": 2185,
-          "clean_text": ""
-        },
-        "author": null,
-        "producer": {
-          "text": "",
-          "confidence": 0.0,
-          "start": 2160,
-          "end": 2185,
-          "clean_text": ""
-        },
-        "geography": {
-          "text": "",
-          "confidence": 0.0,
-          "start": 2160,
-          "end": 2185,
-          "clean_text": ""
-        },
-        "publication_year": null,
-        "reference_year": {
-          "text": "",
-          "confidence": 0.0,
-          "start": 2160,
-          "end": 2185
-        },
-        "reference_population": null,
-        "is_used": "False",
-        "usage_context": "background",
-        "validations": []
-      },
       {
         "dataset_name": {
           "text": "Host country nationals (2023)",
@@ -619,51 +484,6 @@
         "usage_context": "background",
         "validations": []
       },
-      {
-        "dataset_name": {
-          "text": "Eurostat’s at-risk-of-poverty (AROP)",
-          "confidence": 0.34223315119743347,
-          "start": 2329,
-          "end": 2365,
-          "clean_text": "Eurostat’s at-risk-of-poverty (AROP)"
-        },
-        "dataset_tag": "descriptive",
-        "description": null,
-        "data_type": "other",
-        "acronym": {
-          "text": "",
-          "confidence": 0.0,
-          "start": 2329,
-          "end": 2365,
-          "clean_text": ""
-        },
-        "author": null,
-        "producer": {
-          "text": "",
-          "confidence": 0.0,
-          "start": 2329,
-          "end": 2365,
-          "clean_text": ""
-        },
-        "geography": {
-          "text": "",
-          "confidence": 0.0,
-          "start": 2329,
-          "end": 2365,
-          "clean_text": ""
-        },
-        "publication_year": null,
-        "reference_year": {
-          "text": "",
-          "confidence": 0.0,
-          "start": 2329,
-          "end": 2365
-        },
-        "reference_population": null,
-        "is_used": "False",
-        "usage_context": "background",
-        "validations": []
-      },
       {
         "dataset_name": {
           "text": "survey data",
@@ -754,51 +574,6 @@
         "usage_context": "background",
         "validations": []
       },
-      {
-        "dataset_name": {
-          "text": "Eurostat",
-          "confidence": 0.8715351819992065,
-          "start": 2735,
-          "end": 2743,
-          "clean_text": "Eurostat"
-        },
-        "dataset_tag": "descriptive",
-        "description": null,
-        "data_type": "other",
-        "acronym": {
-          "text": "",
-          "confidence": 0.0,
-          "start": 2735,
-          "end": 2743,
-          "clean_text": ""
-        },
-        "author": null,
-        "producer": {
-          "text": "",
-          "confidence": 0.0,
-          "start": 2735,
-          "end": 2743,
-          "clean_text": ""
-        },
-        "geography": {
-          "text": "",
-          "confidence": 0.0,
-          "start": 2735,
-          "end": 2743,
-          "clean_text": ""
-        },
-        "publication_year": null,
-        "reference_year": {
-          "text": "",
-          "confidence": 0.0,
-          "start": 2735,
-          "end": 2743
-        },
-        "reference_population": null,
-        "is_used": "True",
-        "usage_context": "supporting",
-        "validations": []
-      },
       {
         "dataset_name": {
           "text": "SAG estimates",
@@ -855,51 +630,6 @@
   {
     "input_text": "5\nHIGH EMPLOYMENT RATES, BUT LOW WAGES: A POVERTY ASSESSMENT OF UKRAINIAN REFUGEES IN NEIGHBORING COUNTRIES\nLiving in a vulnerable household is associated with higher poverty risks \nJust like last year, members of households with vulnerabilities were found to more likely be living in poverty \nthan the general refugee population. Almost half of individuals living with an older adult (age 65+) reported an \nequivalized disposable income below the poverty threshold. For individuals living with household members \nwith a disability or members with MHPSS needs11 these rates stood at 41% and 27%, respectively. Gender of the \nhead of household, however, was not found to have a significant impact on the poverty rate. Mixed gender (at \nleast one male and one female head) was associated with higher income, though likely due to increased \nchances of multiple breadwinners being present in the household. \nREFUGEE POVERTY RATES BY VULNERABILITY CHARACTERISTIC AND GENDER OF HEAD OF HOUSEHOLD\nOlder adult\n(65+) present\nHousehold\nmember with\na disability\npresent\nHousehold\nmember with\nMHPSS needs\npresent\n18%\n20%\n17%\n47%\n41%\n27%\nNo\nYes\nHead of\nhousehold\ngender\n25%\n24%\n18%\nFemale\nMale\nMixed\nSource: Survey data, SAG estimates\nRefugee housing expenses are on average much higher than for nationals, which implies an \neven greater disparity in financial wellbeing \nAt the regional level, the weighted average share of the host population living in rented housing was calculated \nat 13% based on Eurostat data. This figure is dwarfed by 60% of refugee households fully paying rent for their \naccommodation and 21% partially paying, as per the SEIS survey. Likewise, accommodation expenses as a \nshare of disposable income were estimated at 17% for hosts, including mortgages, compared to 32% for \nrefugees. This essentially implies that refugees, on average, have a much smaller share of their disposable \nincome remaining for other needs. To estimate the magnitude of this effect, poverty rates were recalculated \nafter subtracting the portion of housing expenses that exceed the national average directly from disposable \nincome. This calculation almost doubled the poverty rate, bringing it to 40%. \n11.\t Defined as someone feeling so upset, anxious, worried, agitated, or depressed that it affects daily functioning\n",
     "datasets": [
-      {
-        "dataset_name": {
-          "text": "POVERTY ASSESSMENT OF UKRAINIAN REFUGEES",
-          "confidence": 0.7182486653327942,
-          "start": 42,
-          "end": 82,
-          "clean_text": "POVERTY ASSESSMENT OF UKRAINIAN REFUGEES"
-        },
-        "dataset_tag": "named",
-        "description": null,
-        "data_type": "other",
-        "acronym": {
-          "text": "",
-          "confidence": 0.9721941947937012,
-          "start": null,
-          "end": null,
-          "clean_text": ""
-        },
-        "author": null,
-        "producer": {
-          "text": "MHPSS",
-          "confidence": 0.423165500164032,
-          "start": 544,
-          "end": 549,
-          "clean_text": "MHPSS"
-        },
-        "geography": {
-          "text": "",
-          "confidence": 0.0,
-          "start": 42,
-          "end": 82,
-          "clean_text": ""
-        },
-        "publication_year": null,
-        "reference_year": {
-          "text": "",
-          "confidence": 0.0,
-          "start": 42,
-          "end": 82
-        },
-        "reference_population": null,
-        "is_used": "True",
-        "usage_context": "primary",
-        "validations": []
-      },
       {
         "dataset_name": {
           "text": "Mixed Source: Survey data",
@@ -1720,51 +1450,6 @@
   {
     "input_text": "10\t\nHIGH EMPLOYMENT RATES, BUT LOW WAGES: A POVERTY ASSESSMENT OF UKRAINIAN REFUGEES IN NEIGHBORING COUNTRIES\nRefugee employment rates significantly improved in 2024, moving closer to national levels\nUkrainian refugee employment rates14 have experienced a sizeable increase from 2023 in most countries, \nrising by 9 percentage points year-over-year at the regional level to 64%. This indicator now stands just 7 \npercentage points below the equivalently weighted mean for the host population (71%). \nRefugee employment grew both owing to a decrease in unemployment, which declined to 9% from 13% last \nyear, as well as new individuals coming into the labor force15. In fact, the 2024 labor force participation rate \namounted to 70%, which stands almost in line with the equivalent host country indicator of 73%.\n14.\t The employment rate is defined as the number of employed or self-employed individuals of working age (15-64) as a share of \nthe total number of people in this age group \n15.\t The labor force is defined as the number of people that are either employed or unemployed\nEVOLUTION OF THE REFUGEE EMPLOYMENT RATE FROM 2023 TO 2024\n0%\n10%\n20%\n30%\n40%\n50%\n60%\n70%\nEmployment\nrate 2023\nDecrease in\nunemployment\nDecrease in\npotential\nlabor force\nExpansion of\nthe labor\nforce\nUnclear\nlabor force\nstatus\nEmployment\nrate 2024\nNote: Only includes data from the 7 countries surveyed in both rounds (Bulgaria, Czech Republic, Hungary, Republic of Moldova, Poland, Romania, and \nSlovakia) \nSource: Survey data, SAG estimates\nREFUGEE VS HOST EMPLOYMENT RATES BY COUNTRY\nBulgaria\nCzechia\nEstonia\nHungary\nLatvia\nLithuania\nMoldova\nPoland\nRomania\nSlovakia\nRegion\n38%\n62%\n64%\n36%\n56%\n31%\n52%\n55%\n57%\n76%\n67%\n60%\n52%\n49%\n44%\n67%\n40%\n65%\n64%\n70%\n73%\n76%\n73%\n73%\n71%\n72%\n70%\n62%\n70%\n71%\nRefugee (2023)\nRefugee (2024)\nHost (2023)\nNote: For comparability, employment rates for host countries have been recalculated assuming a similar gender distribution to that of refugees  \nSource: ILO, survey data\n",
     "datasets": [
-      {
-        "dataset_name": {
-          "text": "POVERTY ASSESSMENT OF UKRAINIAN REFUGEES",
-          "confidence": 0.5256072282791138,
-          "start": 43,
-          "end": 83,
-          "clean_text": "POVERTY ASSESSMENT OF UKRAINIAN REFUGEES"
-        },
-        "dataset_tag": "named",
-        "description": null,
-        "data_type": "other",
-        "acronym": {
-          "text": "",
-          "confidence": 0.0,
-          "start": 43,
-          "end": 83,
-          "clean_text": ""
-        },
-        "author": null,
-        "producer": {
-          "text": "",
-          "confidence": 0.0,
-          "start": 43,
-          "end": 83,
-          "clean_text": ""
-        },
-        "geography": {
-          "text": "",
-          "confidence": 0.0,
-          "start": 43,
-          "end": 83,
-          "clean_text": ""
-        },
-        "publication_year": null,
-        "reference_year": {
-          "text": "2024",
-          "confidence": 0.7183356285095215,
-          "start": 674,
-          "end": 678
-        },
-        "reference_population": null,
-        "is_used": "True",
-        "usage_context": "primary",
-        "validations": []
-      },
       {
         "dataset_name": {
           "text": ": Survey data",
@@ -1812,126 +1497,36 @@
       },
       {
         "dataset_name": {
-          "text": "Refugee (2023)",
-          "confidence": 0.5237794518470764,
-          "start": 1767,
-          "end": 1781,
-          "clean_text": "Refugee (2023)"
+          "text": "ILO, survey data",
+          "confidence": 0.6461921334266663,
+          "start": 1960,
+          "end": 1976,
+          "clean_text": "ILO, survey data"
         },
         "dataset_tag": "named",
         "description": null,
         "data_type": "other",
         "acronym": {
           "text": "",
-          "confidence": 0.0,
-          "start": 1767,
-          "end": 1781,
+          "confidence": 0.9847477078437805,
+          "start": null,
+          "end": null,
           "clean_text": ""
         },
         "author": null,
         "producer": {
-          "text": "",
-          "confidence": 0.0,
-          "start": 1767,
-          "end": 1781,
-          "clean_text": ""
+          "text": "ILO",
+          "confidence": 0.9719426035881042,
+          "start": 1960,
+          "end": 1963,
+          "clean_text": "ILO"
         },
         "geography": {
-          "text": "",
-          "confidence": 0.0,
-          "start": 1767,
-          "end": 1781,
-          "clean_text": ""
-        },
-        "publication_year": null,
-        "reference_year": {
-          "text": "2024",
-          "confidence": 0.6904320120811462,
-          "start": 1791,
-          "end": 1795
-        },
-        "reference_population": null,
-        "is_used": "True",
-        "usage_context": "supporting",
-        "validations": []
-      },
-      {
-        "dataset_name": {
-          "text": "Refugee (2024)",
-          "confidence": 0.4031795263290405,
-          "start": 1782,
-          "end": 1796,
-          "clean_text": "Refugee (2024)"
-        },
-        "dataset_tag": "named",
-        "description": null,
-        "data_type": "other",
-        "acronym": {
-          "text": "",
-          "confidence": 0.0,
-          "start": 1782,
-          "end": 1796,
-          "clean_text": ""
-        },
-        "author": null,
-        "producer": {
-          "text": "",
-          "confidence": 0.0,
-          "start": 1782,
-          "end": 1796,
-          "clean_text": ""
-        },
-        "geography": {
-          "text": "",
-          "confidence": 0.0,
-          "start": 1782,
-          "end": 1796,
-          "clean_text": ""
-        },
-        "publication_year": null,
-        "reference_year": {
-          "text": "",
-          "confidence": 0.0,
-          "start": 1782,
-          "end": 1796
-        },
-        "reference_population": null,
-        "is_used": "True",
-        "usage_context": "primary",
-        "validations": []
-      },
-      {
-        "dataset_name": {
-          "text": "ILO, survey data",
-          "confidence": 0.6461921334266663,
-          "start": 1960,
-          "end": 1976,
-          "clean_text": "ILO, survey data"
-        },
-        "dataset_tag": "named",
-        "description": null,
-        "data_type": "other",
-        "acronym": {
-          "text": "",
-          "confidence": 0.9847477078437805,
-          "start": null,
-          "end": null,
-          "clean_text": ""
-        },
-        "author": null,
-        "producer": {
-          "text": "ILO",
-          "confidence": 0.9719426035881042,
-          "start": 1960,
-          "end": 1963,
-          "clean_text": "ILO"
-        },
-        "geography": {
-          "text": "Slovakia",
-          "confidence": 0.9999076128005981,
-          "start": 1631,
-          "end": 1639,
-          "clean_text": "Slovakia"
+          "text": "Slovakia",
+          "confidence": 0.9999076128005981,
+          "start": 1631,
+          "end": 1639,
+          "clean_text": "Slovakia"
         },
         "publication_year": null,
         "reference_year": {
@@ -1956,51 +1551,6 @@
   {
     "input_text": "11\nHIGH EMPLOYMENT RATES, BUT LOW WAGES: A POVERTY ASSESSMENT OF UKRAINIAN REFUGEES IN NEIGHBORING COUNTRIES\nJust like last year, employment rates demonstrated disparity by age group, gender, education level, the \npresence of a disability, and being in need of MHPSS. Disability was found to be associated with the largest \ndrop in employment, with this factor halving the probability of working in 2024 when compared to the sample \noverall. Having MHPSS needs and being female were also associated with a lower employment likelihood, \nalthough to significantly smaller degrees. \nCompared to the host population, the refugee sample demonstrated higher employment rates at lower age \nbrackets (15-19 and 20-24), which, considering the findings on poverty, can be interpreted as a coping strategy \nassociated with low income. Higher age brackets (55-59 and 60-64) saw the highest gap in the employment \nrate compared to hosts, supporting the hypothesis that older age individuals have more difficulty integrating \ninto the local labor market. \nREGIONAL EMPLOYMENT RATE BY AGE AND POPULATION\n15-19\n20-24\n25-29\n30-34\n35-39\n40-44\n45-49\n50-54\n55-59\n60-64\n20%\n64%\n65%\n58%\n62%\n67%\n72%\n61%\n46%\n26%\n25%\n72%\n69%\n73%\n75%\n75%\n77%\n73%\n60%\n36%\n7%\n49%\n79%\n82%\n85%\n87%\n88%\n85%\n80%\n52%\nRefugees (2023)\nRefugees (2024)\nHosts (2023)\nSource: Survey data, ILO\nREGIONAL REFUGEE EMPLOYMENT RATE BY POPULATION \nGROUP (2024)\nOverall\nMale\nFemale\nWith severe psychological\ndistress\nWith a disability\n64%\n67%\n63%\n57%\n30%\nSource: Survey data, SAG estimates \nEMPLOYMENT RATE BY EDUCATION LEVEL\nLower\nsecondary or\nbelow\nTechnical or\nVocational\nBachelor's\nMaster's\nDoctoral\n49%\n59%\n59%\n63%\n54%\n40%\n70%\n72%\n72%\n71%\n37%\n72%\n86%\n92%\n96%\nRefugees (2023)\nRefugees (2024)\nHosts (2023)\nSource: Survey data, SAG estimates\n",
     "datasets": [
-      {
-        "dataset_name": {
-          "text": "POVERTY ASSESSMENT OF UKRAINIAN REFUGEES",
-          "confidence": 0.739567220211029,
-          "start": 43,
-          "end": 83,
-          "clean_text": "POVERTY ASSESSMENT OF UKRAINIAN REFUGEES"
-        },
-        "dataset_tag": "named",
-        "description": null,
-        "data_type": "other",
-        "acronym": {
-          "text": "",
-          "confidence": 0.9873629808425903,
-          "start": null,
-          "end": null,
-          "clean_text": ""
-        },
-        "author": null,
-        "producer": {
-          "text": "MHPSS",
-          "confidence": 0.5941052436828613,
-          "start": 260,
-          "end": 265,
-          "clean_text": "MHPSS"
-        },
-        "geography": {
-          "text": "UKRAINIAN REFUGEES",
-          "confidence": 0.40661299228668213,
-          "start": 65,
-          "end": 83,
-          "clean_text": "UKRAINIAN REFUGEES"
-        },
-        "publication_year": null,
-        "reference_year": {
-          "text": "",
-          "confidence": 0.0,
-          "start": 43,
-          "end": 83
-        },
-        "reference_population": null,
-        "is_used": "True",
-        "usage_context": "primary",
-        "validations": []
-      },
       {
         "dataset_name": {
           "text": ": Survey data",
@@ -2091,51 +1641,6 @@
         "usage_context": "supporting",
         "validations": []
       },
-      {
-        "dataset_name": {
-          "text": "Refugees (2023)",
-          "confidence": 0.3293042480945587,
-          "start": 1689,
-          "end": 1704,
-          "clean_text": "Refugees (2023)"
-        },
-        "dataset_tag": "descriptive",
-        "description": null,
-        "data_type": "other",
-        "acronym": {
-          "text": "",
-          "confidence": 0.0,
-          "start": 1689,
-          "end": 1704,
-          "clean_text": ""
-        },
-        "author": null,
-        "producer": {
-          "text": "",
-          "confidence": 0.0,
-          "start": 1689,
-          "end": 1704,
-          "clean_text": ""
-        },
-        "geography": {
-          "text": "",
-          "confidence": 0.0,
-          "start": 1689,
-          "end": 1704,
-          "clean_text": ""
-        },
-        "publication_year": null,
-        "reference_year": {
-          "text": "",
-          "confidence": 0.0,
-          "start": 1689,
-          "end": 1704
-        },
-        "reference_population": null,
-        "is_used": "False",
-        "usage_context": "background",
-        "validations": []
-      },
       {
         "dataset_name": {
           "text": "Survey data",
@@ -2338,51 +1843,6 @@
   {
     "input_text": "13\nHIGH EMPLOYMENT RATES, BUT LOW WAGES: A POVERTY ASSESSMENT OF UKRAINIAN REFUGEES IN NEIGHBORING COUNTRIES\nUKRAINE REFUGEE VS HOST GROSS MONTHLY WAGES, EUR/MONTH\n18. The difference in median wages by highest education level attained. Weighted equivalently to refugee weights for comparability\n19. The wage gap for refugees was computed based on the difference in weighted means (instead of the difference in medians)\n0\n500\n1,000\n1,500\n2,000\n2,500\nBulgaria\nCzechia\nEstonia\nHungary\nLatvia\nLithuania\nMoldova\nPoland\nRomania\nSlovakia\nRegion\nMinimum wage (2024)\nRefugee mean wage (2023)\nRefugee mean wage (2024)\nHost mean wage (2024)\nNote: Mean wages have been estimated by dividing household employment income by the total number of working hours and then computing a \nweighted average across households with weights proportional to total working hours. As the survey asked for net income, weighted means were then \nconverted to gross amounts for comparability based on host country tax rates.\nSource: Survey data, Eurostat, SAG estimates\nSimilar to its impact on employment status, education seems to have a much less pronounced effect on wage \npremiums for refugees compared to hosts, also suggesting the presence of underemployment. While, \naccording to Eurostat data and SAG estimates, a local with an advanced degree can expect to earn nearly 80% \nmore than someone with only lower secondary education18, the same wage gap19 for Ukrainians stands as just \n16% based on survey data. \nSkills-job mismatching also becomes evident when analyzing the current employment of refugees compared to \ntheir pre-war employment in Ukraine, as nearly 60% have transitioned to entirely different economic sectors. \nThis phenomenon is more pronounced among women, with 63% having shifted to roles outside their previous \nemployment background, compared to 50% of men. One possible explanation for this discrepancy is the higher \nproportion of men employed in sectors like construction and IT prior to displacement. These fields often \ndemand fewer country-specific qualifications, such as proficiency in the local language, thereby facilitating \neasier integration into similar roles in the host country.\nDISTRIBUTION OF WORKING AGE POPULATION BY \nHIGHEST EDUCATION LEVEL ATTAINED\nHosts (2023)\nRefugees (2024)\n−56%\n−24%\n9%\n12%\n18%\n39%\nTechnical or Vocational\nBachelor\nMaster's\nDoctoral\n1%\n1%\nSource: Survey data, ILO\nWAGE PREMIUMS FOR EDUCATION LEVEL GAINS: HOSTS \nVERSUS REFUGEES, %\nTechnical or\nvocational\nBachelor's or\nabove\n15%\n78%\n3%\n16%\nHost education wage\npremium (median,\n2022)\nRefugee eduction\nwage premium\n(mean, 2024)\nNote: Wage premiums computed with lower secondary education as \nthe baseline\nSource: Survey data, Eurostat, SAG estimates\n",
     "datasets": [
-      {
-        "dataset_name": {
-          "text": "Region Minimum wage (2024)",
-          "confidence": 0.3194044232368469,
-          "start": 531,
-          "end": 557,
-          "clean_text": "Region Minimum wage (2024)"
-        },
-        "dataset_tag": "descriptive",
-        "description": null,
-        "data_type": "other",
-        "acronym": {
-          "text": "",
-          "confidence": 0.0,
-          "start": 531,
-          "end": 557,
-          "clean_text": ""
-        },
-        "author": null,
-        "producer": {
-          "text": "",
-          "confidence": 0.0,
-          "start": 531,
-          "end": 557,
-          "clean_text": ""
-        },
-        "geography": {
-          "text": "",
-          "confidence": 0.0,
-          "start": 531,
-          "end": 557,
-          "clean_text": ""
-        },
-        "publication_year": null,
-        "reference_year": {
-          "text": "",
-          "confidence": 0.0,
-          "start": 531,
-          "end": 557
-        },
-        "reference_population": null,
-        "is_used": "False",
-        "usage_context": "background",
-        "validations": []
-      },
       {
         "dataset_name": {
           "text": "Refugee mean wage (2023)",
@@ -2563,51 +2023,6 @@
         "usage_context": "supporting",
         "validations": []
       },
-      {
-        "dataset_name": {
-          "text": "Eurostat",
-          "confidence": 0.9324009418487549,
-          "start": 1250,
-          "end": 1258,
-          "clean_text": "Eurostat"
-        },
-        "dataset_tag": "named",
-        "description": null,
-        "data_type": "estimates",
-        "acronym": {
-          "text": "",
-          "confidence": 0.0,
-          "start": 1250,
-          "end": 1258,
-          "clean_text": ""
-        },
-        "author": null,
-        "producer": {
-          "text": "",
-          "confidence": 0.0,
-          "start": 1250,
-          "end": 1258,
-          "clean_text": ""
-        },
-        "geography": {
-          "text": "",
-          "confidence": 0.0,
-          "start": 1250,
-          "end": 1258,
-          "clean_text": ""
-        },
-        "publication_year": null,
-        "reference_year": {
-          "text": "",
-          "confidence": 0.0,
-          "start": 1250,
-          "end": 1258
-        },
-        "reference_population": null,
-        "is_used": "True",
-        "usage_context": "supporting",
-        "validations": []
-      },
       {
         "dataset_name": {
           "text": "survey data",
@@ -2697,51 +2112,6 @@
         "is_used": "True",
         "usage_context": "primary",
         "validations": []
-      },
-      {
-        "dataset_name": {
-          "text": "ILO WAGE PREMIUMS FOR EDUCATION LEVEL GAINS",
-          "confidence": 0.32336169481277466,
-          "start": 2385,
-          "end": 2428,
-          "clean_text": "ILO WAGE PREMIUMS FOR EDUCATION LEVEL GAINS"
-        },
-        "dataset_tag": "descriptive",
-        "description": null,
-        "data_type": "other",
-        "acronym": {
-          "text": "",
-          "confidence": 0.0,
-          "start": 2385,
-          "end": 2428,
-          "clean_text": ""
-        },
-        "author": null,
-        "producer": {
-          "text": "",
-          "confidence": 0.0,
-          "start": 2385,
-          "end": 2428,
-          "clean_text": ""
-        },
-        "geography": {
-          "text": "",
-          "confidence": 0.0,
-          "start": 2385,
-          "end": 2428,
-          "clean_text": ""
-        },
-        "publication_year": null,
-        "reference_year": {
-          "text": "",
-          "confidence": 0.0,
-          "start": 2385,
-          "end": 2428
-        },
-        "reference_population": null,
-        "is_used": "True",
-        "usage_context": "primary",
-        "validations": []
       }
     ],
     "document": {
@@ -2844,51 +2214,6 @@
         "usage_context": "primary",
         "validations": []
       },
-      {
-        "dataset_name": {
-          "text": "Eurostat",
-          "confidence": 0.8321400284767151,
-          "start": 2513,
-          "end": 2521,
-          "clean_text": "Eurostat"
-        },
-        "dataset_tag": "named",
-        "description": null,
-        "data_type": "other",
-        "acronym": {
-          "text": "",
-          "confidence": 0.6210211515426636,
-          "start": null,
-          "end": null,
-          "clean_text": ""
-        },
-        "author": null,
-        "producer": {
-          "text": "",
-          "confidence": 0.0,
-          "start": 2513,
-          "end": 2521,
-          "clean_text": ""
-        },
-        "geography": {
-          "text": "",
-          "confidence": 0.0,
-          "start": 2513,
-          "end": 2521,
-          "clean_text": ""
-        },
-        "publication_year": null,
-        "reference_year": {
-          "text": "2023",
-          "confidence": 0.78436678647995,
-          "start": 2526,
-          "end": 2530
-        },
-        "reference_population": null,
-        "is_used": "True",
-        "usage_context": "primary",
-        "validations": []
-      },
       {
         "dataset_name": {
           "text": "national annual wage inflation data",
@@ -3125,51 +2450,6 @@
         "usage_context": "primary",
         "validations": []
       },
-      {
-        "dataset_name": {
-          "text": "EU SILC",
-          "confidence": 0.5524539351463318,
-          "start": 2054,
-          "end": 2061,
-          "clean_text": "EU SILC"
-        },
-        "dataset_tag": "named",
-        "description": null,
-        "data_type": "other",
-        "acronym": {
-          "text": "",
-          "confidence": 0.0,
-          "start": 2054,
-          "end": 2061,
-          "clean_text": ""
-        },
-        "author": null,
-        "producer": {
-          "text": "",
-          "confidence": 0.0,
-          "start": 2054,
-          "end": 2061,
-          "clean_text": ""
-        },
-        "geography": {
-          "text": "",
-          "confidence": 0.0,
-          "start": 2054,
-          "end": 2061,
-          "clean_text": ""
-        },
-        "publication_year": null,
-        "reference_year": {
-          "text": "",
-          "confidence": 0.0,
-          "start": 2054,
-          "end": 2061
-        },
-        "reference_population": null,
-        "is_used": "True",
-        "usage_context": "supporting",
-        "validations": []
-      },
       {
         "dataset_name": {
           "text": "Republic of Moldova’s Household Budget Survey",
@@ -3225,53 +2505,7 @@
   },
   {
     "input_text": "March 2025\nHIGH EMPLOYMENT \nRATES, BUT LOW \nWAGES: A POVERTY \nASSESSMENT OF \nUKRAINIAN REFUGEES \nIN NEIGHBORING \nCOUNTRIES \nAn inter-agency exploration of \nsocio-economic data\n",
-    "datasets": [
-      {
-        "dataset_name": {
-          "text": "POVERTY ASSESSMENT OF UKRAINIAN REFUGEES IN NEIGHBORING COUNTRIES",
-          "confidence": 0.43391868472099304,
-          "start": 51,
-          "end": 116,
-          "clean_text": "POVERTY ASSESSMENT OF UKRAINIAN REFUGEES IN NEIGHBORING COUNTRIES"
-        },
-        "dataset_tag": "named",
-        "description": null,
-        "data_type": "other",
-        "acronym": {
-          "text": "",
-          "confidence": 0.0,
-          "start": 51,
-          "end": 116,
-          "clean_text": ""
-        },
-        "author": null,
-        "producer": {
-          "text": "",
-          "confidence": 0.0,
-          "start": 51,
-          "end": 116,
-          "clean_text": ""
-        },
-        "geography": {
-          "text": "",
-          "confidence": 0.0,
-          "start": 51,
-          "end": 116,
-          "clean_text": ""
-        },
-        "publication_year": null,
-        "reference_year": {
-          "text": "March 2025",
-          "confidence": 0.7396136522293091,
-          "start": 0,
-          "end": 10
-        },
-        "reference_population": null,
-        "is_used": "True",
-        "usage_context": "primary",
-        "validations": []
-      }
-    ],
+    "datasets": [],
     "document": {
       "source": "/sample_pdfs/2025 03  SEIS ukraine_socio-economic_researchpaper.pdf",
       "pages": [

annotation_data/sample_extractions/doc_3/raw/doc_3_direct_judged.jsonl

@@ -618,51 +618,6 @@
         "usage_context": "background",
         "validations": []
       },
-      {
-        "dataset_name": {
-          "text": "PESEL UKR",
-          "confidence": 0.44700488448143005,
-          "start": 3508,
-          "end": 3517,
-          "clean_text": "PESEL UKR"
-        },
-        "dataset_tag": "named",
-        "description": null,
-        "data_type": "other",
-        "acronym": {
-          "text": "",
-          "confidence": 0.0,
-          "start": 3508,
-          "end": 3517,
-          "clean_text": ""
-        },
-        "author": null,
-        "producer": {
-          "text": "",
-          "confidence": 0.0,
-          "start": 3508,
-          "end": 3517,
-          "clean_text": ""
-        },
-        "geography": {
-          "text": "",
-          "confidence": 0.0,
-          "start": 3508,
-          "end": 3517,
-          "clean_text": ""
-        },
-        "publication_year": null,
-        "reference_year": {
-          "text": "",
-          "confidence": 0.0,
-          "start": 3508,
-          "end": 3517
-        },
-        "reference_population": null,
-        "is_used": "False",
-        "usage_context": "background",
-        "validations": []
-      },
       {
         "dataset_name": {
           "text": "Ukrainians registered for social insurance",
@@ -1641,51 +1596,6 @@
         "usage_context": "background",
         "validations": []
       },
-      {
-        "dataset_name": {
-          "text": "SEIS (May-June 2024",
-          "confidence": 0.3536454141139984,
-          "start": 3717,
-          "end": 3736,
-          "clean_text": "SEIS (May-June 2024"
-        },
-        "dataset_tag": "named",
-        "description": null,
-        "data_type": "other",
-        "acronym": {
-          "text": "",
-          "confidence": 0.0,
-          "start": 3717,
-          "end": 3736,
-          "clean_text": ""
-        },
-        "author": null,
-        "producer": {
-          "text": "",
-          "confidence": 0.0,
-          "start": 3717,
-          "end": 3736,
-          "clean_text": ""
-        },
-        "geography": {
-          "text": "",
-          "confidence": 0.0,
-          "start": 3717,
-          "end": 3736,
-          "clean_text": ""
-        },
-        "publication_year": null,
-        "reference_year": {
-          "text": "",
-          "confidence": 0.0,
-          "start": 3717,
-          "end": 3736
-        },
-        "reference_population": null,
-        "is_used": "False",
-        "usage_context": "background",
-        "validations": []
-      },
       {
         "dataset_name": {
           "text": "UNHCR (2024, 2023) surveys",
@@ -1731,51 +1641,6 @@
         "usage_context": "background",
         "validations": []
       },
-      {
-        "dataset_name": {
-          "text": "Tokariuk, 2025",
-          "confidence": 0.5730606317520142,
-          "start": 4282,
-          "end": 4296,
-          "clean_text": "Tokariuk, 2025"
-        },
-        "dataset_tag": "vague",
-        "description": null,
-        "data_type": "other",
-        "acronym": {
-          "text": "",
-          "confidence": 0.0,
-          "start": 4282,
-          "end": 4296,
-          "clean_text": ""
-        },
-        "author": null,
-        "producer": {
-          "text": "",
-          "confidence": 0.0,
-          "start": 4282,
-          "end": 4296,
-          "clean_text": ""
-        },
-        "geography": {
-          "text": "",
-          "confidence": 0.0,
-          "start": 4282,
-          "end": 4296,
-          "clean_text": ""
-        },
-        "publication_year": null,
-        "reference_year": {
-          "text": "",
-          "confidence": 0.0,
-          "start": 4282,
-          "end": 4296
-        },
-        "reference_population": null,
-        "is_used": "False",
-        "usage_context": "background",
-        "validations": []
-      },
       {
         "dataset_name": {
           "text": "UNHCR",
@@ -2382,51 +2247,6 @@
         "is_used": "True",
         "usage_context": "primary",
         "validations": []
-      },
-      {
-        "dataset_name": {
-          "text": "PESEL UKR",
-          "confidence": 0.30657365918159485,
-          "start": 3864,
-          "end": 3873,
-          "clean_text": "PESEL UKR"
-        },
-        "dataset_tag": "descriptive",
-        "description": null,
-        "data_type": "other",
-        "acronym": {
-          "text": "",
-          "confidence": 0.0,
-          "start": 3864,
-          "end": 3873,
-          "clean_text": ""
-        },
-        "author": null,
-        "producer": {
-          "text": "",
-          "confidence": 0.0,
-          "start": 3864,
-          "end": 3873,
-          "clean_text": ""
-        },
-        "geography": {
-          "text": "",
-          "confidence": 0.0,
-          "start": 3864,
-          "end": 3873,
-          "clean_text": ""
-        },
-        "publication_year": null,
-        "reference_year": {
-          "text": "",
-          "confidence": 0.0,
-          "start": 3864,
-          "end": 3873
-        },
-        "reference_population": null,
-        "is_used": "True",
-        "usage_context": "primary",
-        "validations": []
       }
     ],
     "document": {
@@ -2439,51 +2259,6 @@
   {
     "input_text": "Analysis of the impact of refugees from Ukraine on the economy of Poland\n19\nAnalysis of the impact of refugees from Ukraine on the economy of Poland\n18\nUkrainian refugees in Poland have \nclearly improved their economic \nsituation over the past year. In \nthe 15-59/64 age group, employment \nrate of Polish citizens stood at 75% in \nQ2 2024 according to Eurostat, slightly \nmore than the 69% for Ukrainian refugees \nin the SEIS 2024 survey and 73% when \nadjusted for a different sex and age \nstructure. In the 15-64 age group, the \nemployment rate of Ukrainian male \nrefugees was 67% while that of Polish \ncitizens in Q2 2024 was 77%. For women, \nthe rates for Ukrainian refugees and Polish \ncitizens are closer in the 15-59 age group \n(female retirement age in Poland is 60) with \n70% for refugees and 72% for Poles. Visible \ndifferences are identified in 15-19 and \n20-24 age groups, with much higher \nemployment rates for Ukrainian refugees \ndue perhaps to the fact that the Ukrainian \nschool system ends at 17 while the Polish \none at 19.13 However, employment rates \ndrop for Ukrainian female refugees in the \n25-29, 30-34, and 35-39 age groups when \ncompared to younger and older groups, as \nwell as to Polish women, which may be due \nto insufficient access to childcare services. \nEmployment rates for Ukrainian female \nrefugees are also substantially lower in the \n55-59 age group, perhaps due to the fact \nthat female retirement age in Ukraine has \nonly recently (in 2021) become 60.14\nWith the current level of integration \nof Ukrainian refugees into the labour \nmarket and the jobs they perform, \ntheir median net earnings are about \nfour-fifths of the national median \n– although likely lower in terms of \naverage or gross earnings. The median \nnet earnings of Ukrainian refugees in \nQ2 2024 were PLN 4,000 in SEIS, and \nPLN 3,767 in NBP (2024) surveys. This \nis 84% and 79% of the national median, \nrespectively. This estimate would be \nmost likely lower, if data allowed us to \nlook at average or gross earnings. First, a \ncomparison of average instead of median \nearnings would show a larger gap between \nUkrainian refugees and all workers in \nthe economy, as the national average is \ninflated by top incomes, which are less \noften earned by Ukrainian refugees than \nthe general population. UNHCR (2025b), \nusing a different method than the one in \nthis report, estimated average instead \nof median net wages based on the SEIS \ndata. The average net wage of a Ukrainian \nrefugee they arrived at was PLN 4,214, only \nslightly higher than calculated above. This \nwould yield 72% of the national average net \nwage.15 Average is not used here, because \nmedian is more relevant for discussing \neconomic impact, as it is not disrupted \nby bottom or top earnings. Furthermore, \nthere are no outside estimates to compare \nit to. Second, as discussed previously, there \nare much fewer Ukrainian refugees with \nemployment contracts than Poles, and \nthis lowers their social contributions and \nthus gross earnings. Unfortunately, there \nis no such data available, as both SEIS and \nNBP (2024) measure only net earnings. \nAdministrative ZUS data can be used as a \nproxy for both average and gross earnings \npercentages. On June 30, 2024, average \nbases for social contributions of Ukrainian \nrefugees accounted for just 64% of those \nof Polish citizens. However, this is only \nrelative to Polish citizens not all workers in \nthe economy as a whole, and comes with \nother caveats of administrative instead \nof survey data sources – data given on a \nparticular day instead of period average, \nthe shadow economy unaccounted for, \ndifferent contributions based on the type \nof contract, no data for farmers who belong \nto a separate social insurance scheme. \nAlso, the ZUS data, that is available, is much \nmore limited that the SEIS survey data \nprimarily used in this report.\n2.2 Current occupational situation\n13\tChildren in Ukraine start school at 6 years of age, while in Poland at 7. School system lasts for 11 years in Ukraine and 12 years in Poland. After recent reform, \nUkrainian children who started school in or after 2018 will receive 12 years of schooling.\n14\thttps://www.social-protection.org/gimi/gess/Media.action;jsessionid=U1dm6XAPF38pPk5_pzhfaGoS_rOACMDDVv4w5uevMsKBeQEC5-_g!284293951?id=15680\n15\tPercentage calculated based on GUS average for the economy as a whole in the months of the SEIS survey.\nSource: Deloitte own elaboration based on Eurostat Labour Force Survey data SEIS UNHCR survey \nconducted in May and June 2024. For Polish citizens reference period is Q2 2024.\nChart 13. Polish citizens and Ukrainian refugees’ employment rates by age group\n0\n0\n20%\n20%\n80%\n80%\n40%\n40%\n100%\n100%\n60%\n60%\nMale\nFemale\n15-19\n15-19\n40-44\n40-44\n20-24\n20-24\n45-49\n45-49\n25-29\n25-29\n50-54\n50-54\n30-34\n30-34\n55-59\n55-59\n35-39\n35-39\n60-64\n60-64\n15-64\n15-59\nPolish citizens\nUkrainian refugees\nChart 14. Ukrainian refugee median net wage estimates in Q2 2024\n0\nMonthly net wage (PLN)\nPercengate of all workers total economy average\nUNHCR\n15 May - 24 June 2024\nUNHCR\n15 May - 24 June 2024\nNBP\n6 May - 5 July 2024\nNBP\n6 May - 5 July 2024\n4,000\n84%\n79%\n3,767\nSource: Deloitte own elaboration based on SEIS survey, NBP (2024) survey. Monthly GUS median wage in the general economy has been recalculated to reflect the \nspecific time periods of SEIS UNHCR and NBP (2024) surveys.\nThe wages of Ukrainian refugees are \nhigher for men than women, but the \ngap is not wider than in the economy \nas a whole. The median net wage of \nfemale Ukrainian refugees is 6% lower than \nin the male group (compared to 7% in the \neconomy as a whole), however, the gap \nmay be different due to the fact that there \nare fewer men in the Ukrainian refugee \npopulation.\n",
     "datasets": [
-      {
-        "dataset_name": {
-          "text": "Eurostat",
-          "confidence": 0.9223920702934265,
-          "start": 347,
-          "end": 355,
-          "clean_text": "Eurostat"
-        },
-        "dataset_tag": "named",
-        "description": null,
-        "data_type": "survey",
-        "acronym": {
-          "text": "",
-          "confidence": 0.7656297087669373,
-          "start": null,
-          "end": null,
-          "clean_text": ""
-        },
-        "author": null,
-        "producer": {
-          "text": "",
-          "confidence": 0.0,
-          "start": 347,
-          "end": 355,
-          "clean_text": ""
-        },
-        "geography": {
-          "text": "Ukraine",
-          "confidence": 0.9795709848403931,
-          "start": 116,
-          "end": 123,
-          "clean_text": "Ukraine"
-        },
-        "publication_year": null,
-        "reference_year": {
-          "text": "",
-          "confidence": 0.0,
-          "start": 347,
-          "end": 355
-        },
-        "reference_population": null,
-        "is_used": "True",
-        "usage_context": "primary",
-        "validations": []
-      },
       {
         "dataset_name": {
           "text": "SEIS 2024 survey",
@@ -2934,51 +2709,6 @@
         "usage_context": "supporting",
         "validations": []
       },
-      {
-        "dataset_name": {
-          "text": "School system",
-          "confidence": 0.34157389402389526,
-          "start": 3899,
-          "end": 3912,
-          "clean_text": "School system"
-        },
-        "dataset_tag": "named",
-        "description": null,
-        "data_type": "other",
-        "acronym": {
-          "text": "",
-          "confidence": 0.0,
-          "start": 3899,
-          "end": 3912,
-          "clean_text": ""
-        },
-        "author": null,
-        "producer": {
-          "text": "",
-          "confidence": 0.0,
-          "start": 3899,
-          "end": 3912,
-          "clean_text": ""
-        },
-        "geography": {
-          "text": "",
-          "confidence": 0.0,
-          "start": 3899,
-          "end": 3912,
-          "clean_text": ""
-        },
-        "publication_year": null,
-        "reference_year": {
-          "text": "",
-          "confidence": 0.0,
-          "start": 3899,
-          "end": 3912
-        },
-        "reference_population": null,
-        "is_used": "True",
-        "usage_context": "primary",
-        "validations": []
-      },
       {
         "dataset_name": {
           "text": "SEIS survey",
@@ -3361,51 +3091,6 @@
         "usage_context": "background",
         "validations": []
       },
-      {
-        "dataset_name": {
-          "text": "Deloitte D.Climate model",
-          "confidence": 0.9816028475761414,
-          "start": 2125,
-          "end": 2149,
-          "clean_text": "Deloitte D.Climate model"
-        },
-        "dataset_tag": "named",
-        "description": null,
-        "data_type": "other",
-        "acronym": {
-          "text": "",
-          "confidence": 0.8939415216445923,
-          "start": null,
-          "end": null,
-          "clean_text": ""
-        },
-        "author": null,
-        "producer": {
-          "text": "",
-          "confidence": 0.0,
-          "start": 2125,
-          "end": 2149,
-          "clean_text": ""
-        },
-        "geography": {
-          "text": "",
-          "confidence": 0.0,
-          "start": 2125,
-          "end": 2149,
-          "clean_text": ""
-        },
-        "publication_year": null,
-        "reference_year": {
-          "text": "2022",
-          "confidence": 0.7519863843917847,
-          "start": 2230,
-          "end": 2234
-        },
-        "reference_population": null,
-        "is_used": "True",
-        "usage_context": "primary",
-        "validations": []
-      },
       {
         "dataset_name": {
           "text": "previous Deloitte (2024) report",
@@ -3498,74 +3183,29 @@
       },
       {
         "dataset_name": {
-          "text": "Deloitte (2024)",
-          "confidence": 0.7153786420822144,
-          "start": 3839,
-          "end": 3854,
-          "clean_text": "Deloitte (2024)"
+          "text": "IMF report",
+          "confidence": 0.9813289642333984,
+          "start": 4709,
+          "end": 4719,
+          "clean_text": "IMF report"
         },
-        "dataset_tag": "descriptive",
+        "dataset_tag": "named",
         "description": null,
-        "data_type": "other",
+        "data_type": "report",
         "acronym": {
           "text": "",
-          "confidence": 0.9984872341156006,
+          "confidence": 0.9994422793388367,
           "start": null,
           "end": null,
           "clean_text": ""
         },
         "author": null,
         "producer": {
-          "text": "",
-          "confidence": 0.0,
-          "start": 3839,
-          "end": 3854,
-          "clean_text": ""
-        },
-        "geography": {
-          "text": "",
-          "confidence": 0.0,
-          "start": 3839,
-          "end": 3854,
-          "clean_text": ""
-        },
-        "publication_year": null,
-        "reference_year": {
-          "text": "",
-          "confidence": 0.0,
-          "start": 3839,
-          "end": 3854
-        },
-        "reference_population": null,
-        "is_used": "False",
-        "usage_context": "background",
-        "validations": []
-      },
-      {
-        "dataset_name": {
-          "text": "IMF report",
-          "confidence": 0.9813289642333984,
-          "start": 4709,
-          "end": 4719,
-          "clean_text": "IMF report"
-        },
-        "dataset_tag": "named",
-        "description": null,
-        "data_type": "report",
-        "acronym": {
-          "text": "",
-          "confidence": 0.9994422793388367,
-          "start": null,
-          "end": null,
-          "clean_text": ""
-        },
-        "author": null,
-        "producer": {
-          "text": "IMF",
-          "confidence": 0.9949954748153687,
-          "start": 4709,
-          "end": 4712,
-          "clean_text": "IMF"
+          "text": "IMF",
+          "confidence": 0.9949954748153687,
+          "start": 4709,
+          "end": 4712,
+          "clean_text": "IMF"
         },
         "geography": {
           "text": "Poland",
@@ -3676,51 +3316,6 @@
         "usage_context": "background",
         "validations": []
       },
-      {
-        "dataset_name": {
-          "text": "Online Technical Appendix",
-          "confidence": 0.7784573435783386,
-          "start": 6285,
-          "end": 6310,
-          "clean_text": "Online Technical Appendix"
-        },
-        "dataset_tag": "named",
-        "description": null,
-        "data_type": "other",
-        "acronym": {
-          "text": "",
-          "confidence": 0.0,
-          "start": 6285,
-          "end": 6310,
-          "clean_text": ""
-        },
-        "author": null,
-        "producer": {
-          "text": "",
-          "confidence": 0.0,
-          "start": 6285,
-          "end": 6310,
-          "clean_text": ""
-        },
-        "geography": {
-          "text": "",
-          "confidence": 0.0,
-          "start": 6285,
-          "end": 6310,
-          "clean_text": ""
-        },
-        "publication_year": null,
-        "reference_year": {
-          "text": "",
-          "confidence": 0.0,
-          "start": 6285,
-          "end": 6310
-        },
-        "reference_population": null,
-        "is_used": "False",
-        "usage_context": "background",
-        "validations": []
-      },
       {
         "dataset_name": {
           "text": "Ministry of Finance GDP growth forecast",
@@ -3956,51 +3551,6 @@
         "is_used": "False",
         "usage_context": "background",
         "validations": []
-      },
-      {
-        "dataset_name": {
-          "text": "Meager & Speckesser (2011)",
-          "confidence": 0.5072301626205444,
-          "start": 3062,
-          "end": 3088,
-          "clean_text": "Meager & Speckesser (2011)"
-        },
-        "dataset_tag": "named",
-        "description": null,
-        "data_type": "other",
-        "acronym": {
-          "text": "",
-          "confidence": 0.0,
-          "start": 3062,
-          "end": 3088,
-          "clean_text": ""
-        },
-        "author": null,
-        "producer": {
-          "text": "",
-          "confidence": 0.0,
-          "start": 3062,
-          "end": 3088,
-          "clean_text": ""
-        },
-        "geography": {
-          "text": "",
-          "confidence": 0.0,
-          "start": 3062,
-          "end": 3088,
-          "clean_text": ""
-        },
-        "publication_year": null,
-        "reference_year": {
-          "text": "",
-          "confidence": 0.0,
-          "start": 3062,
-          "end": 3088
-        },
-        "reference_population": null,
-        "is_used": "False",
-        "usage_context": "background",
-        "validations": []
       }
     ],
     "document": {
@@ -4013,51 +3563,6 @@
   {
     "input_text": "Analysis of the impact of refugees from Ukraine on the economy of Poland\n27\nAnalysis of the impact of refugees from Ukraine on the economy of Poland\n26\n4.\nUkrainian refugees’ \nskill-job mismatch\nOccupational downgrading affects \nall types of migrants across the \nglobe, including refugees, and has \nlong concerned economists and \npolicymakers. According to a previous \nresearch, usually about a third of migrants \nholding a university degree work in jobs \nrequiring only high-school diplomas, versus \n10% of natives (Tani, 2020). Migrants may \nlack language fluency or country-specific \nknowledge (e.g. of the law or business \ncontacts) and find it difficult to prove the \nprofessional experience or educational \ncredentials gained in their home country. \nThey also face regulatory barriers to \nentering some professions, e.g., in the \npublic sector or regulated specialist \noccupations. Ensuring that migrants` \nand refugees` skills are fully utilized to \nthe benefit of both the individuals and \nhost-country economies is a central policy \nchallenge for the countries facing large-\nscale immigration.\nIn Poland, non-EU27 citizens are twice \nas likely to be over-qualified (for their \njob) as Polish citizens. In its Labour \nForce Survey (LFS), Eurostat defines over-\nqualification rate as the share of persons \nwith tertiary education (bachelor’s degree \nand higher) who are employed in the ISCO \n(International Standard Classification of \nOccupations) 4-9 occupational groups25. \nWhile the LFS is unlikely to cover all \nUkrainian refugees and migrants from \nother nationalities in Poland, there was a \nvisible increase in over-qualification rates \nfor non-EU27 citizens between 2022 and \n2023 that is not observed for Polish \ncitizens.\nOccupational downgrading is \nwidespread among Ukrainian refugees \nin Poland. SEIS data shows that 40% of \nUkrainian refugees aged 25–64 hold a \ntertiary degree – exceeding the share of \nPolish citizens in the same age bracket, \naccording to Eurostat’s 2023 annual \naverage. While ZUS data available as of \nJune 30, 2024, shows occupational groups \nfor less than half of the socially insured \nin Poland, it reflects a visible mismatch \nbetween the refugees’ education and the \njobs they perform. Only 12% of Ukrainian \nrefugees (less than one-third of the share \nin terms of tertiary education) worked in \noccupational groups which require tertiary \neducation, that is managers, specialists, \nand technicians (ISCO 1-3), compared to \n37% of Polish citizens (almost the same \nas the tertiary education share in the \n25-64 age group).\nThe disparity between education levels and the jobs performed \nby Ukrainian refugees indicates their significant over-qualification \nin comparison to Polish citizens (4.1). Widespread occupational \nlicensing presents an obstacle to efficient use of Ukrainian refugees’ \nhuman capital, who – in comparison to Polish citizens – rarely work \nin regulated professions (4.2). The higher the level of Polish language \ncompetence, the higher the wage premium. Older refugees, those \nwho have been in Poland for a shorter period of time or have below \ntertiary education tend to have poorer language skills (4.3). Refugees \nwho back in Ukraine managed the household or were not employed, \nremain underemployed in Poland (4.4). Key factors to enhancing \nthe status of Ukrainian refugees in Poland will be language courses, \nincluding advanced levels tailored for professionals, easing of \noccupational licensing restrictions, particularly those based solely on \ncitizenship, and implementation of active labour market policies to \nhelp specific groups enter the workforce (4.5).\n4.1 Over-qualification\nChart 18. Over-qualification rates by citizenship\nNon-EU27 Citizenship\nPolish Citizenship\nSource: Deloitte own elaboration based on Eurostat \n(Labour Force Survey) data.\n2022\n2023\n40%\n48%\n19%\n20%\n25\tThis is the same classification as quoted from GUS and ZUS in the previous chapters. Codes 4-9 refer to clerical support workers; service and sales workers; skilled \nagricultural, forestry and fishery workers; craft and related trades workers; plant and machine operators and assemblers; elementary occupations.\n",
     "datasets": [
-      {
-        "dataset_name": {
-          "text": "Tani",
-          "confidence": 0.6502283215522766,
-          "start": 508,
-          "end": 512,
-          "clean_text": "Tani"
-        },
-        "dataset_tag": "named",
-        "description": null,
-        "data_type": "other",
-        "acronym": {
-          "text": "",
-          "confidence": 0.996781587600708,
-          "start": null,
-          "end": null,
-          "clean_text": ""
-        },
-        "author": null,
-        "producer": {
-          "text": "",
-          "confidence": 0.0,
-          "start": 508,
-          "end": 512,
-          "clean_text": ""
-        },
-        "geography": {
-          "text": "Ukraine",
-          "confidence": 0.9797425866127014,
-          "start": 40,
-          "end": 47,
-          "clean_text": "Ukraine"
-        },
-        "publication_year": null,
-        "reference_year": {
-          "text": "2020",
-          "confidence": 0.9967557787895203,
-          "start": 514,
-          "end": 518
-        },
-        "reference_population": null,
-        "is_used": "False",
-        "usage_context": "background",
-        "validations": []
-      },
       {
         "dataset_name": {
           "text": "Labour Force Survey",
@@ -4237,51 +3742,6 @@
         "is_used": "True",
         "usage_context": "primary",
         "validations": []
-      },
-      {
-        "dataset_name": {
-          "text": "Eurostat",
-          "confidence": 0.4899872839450836,
-          "start": 3721,
-          "end": 3729,
-          "clean_text": "Eurostat"
-        },
-        "dataset_tag": "named",
-        "description": null,
-        "data_type": "survey",
-        "acronym": {
-          "text": "",
-          "confidence": 0.8332523703575134,
-          "start": null,
-          "end": null,
-          "clean_text": ""
-        },
-        "author": null,
-        "producer": {
-          "text": "Deloitte",
-          "confidence": 0.7880746126174927,
-          "start": 3687,
-          "end": 3695,
-          "clean_text": "Deloitte"
-        },
-        "geography": {
-          "text": "",
-          "confidence": 0.0,
-          "start": 3721,
-          "end": 3729,
-          "clean_text": ""
-        },
-        "publication_year": null,
-        "reference_year": {
-          "text": "2022",
-          "confidence": 0.6172484755516052,
-          "start": 3758,
-          "end": 3762
-        },
-        "reference_population": null,
-        "is_used": "True",
-        "usage_context": "primary",
-        "validations": []
       }
     ],
     "document": {
@@ -4845,51 +4305,6 @@
         "usage_context": "primary",
         "validations": []
       },
-      {
-        "dataset_name": {
-          "text": "ZUS-registered",
-          "confidence": 0.5483613014221191,
-          "start": 2871,
-          "end": 2885,
-          "clean_text": "ZUS-registered"
-        },
-        "dataset_tag": "named",
-        "description": null,
-        "data_type": "other",
-        "acronym": {
-          "text": "",
-          "confidence": 0.0,
-          "start": 2871,
-          "end": 2885,
-          "clean_text": ""
-        },
-        "author": null,
-        "producer": {
-          "text": "",
-          "confidence": 0.0,
-          "start": 2871,
-          "end": 2885,
-          "clean_text": ""
-        },
-        "geography": {
-          "text": "",
-          "confidence": 0.0,
-          "start": 2871,
-          "end": 2885,
-          "clean_text": ""
-        },
-        "publication_year": null,
-        "reference_year": {
-          "text": "",
-          "confidence": 0.0,
-          "start": 2871,
-          "end": 2885
-        },
-        "reference_population": null,
-        "is_used": "True",
-        "usage_context": "supporting",
-        "validations": []
-      },
       {
         "dataset_name": {
           "text": "Babakova et al. (2024) report",
@@ -5643,51 +5058,6 @@
         "usage_context": "background",
         "validations": []
       },
-      {
-        "dataset_name": {
-          "text": "Peri, 2016",
-          "confidence": 0.7205634713172913,
-          "start": 1447,
-          "end": 1457,
-          "clean_text": "Peri, 2016"
-        },
-        "dataset_tag": "named",
-        "description": null,
-        "data_type": "other",
-        "acronym": {
-          "text": "",
-          "confidence": 0.9601743221282959,
-          "start": null,
-          "end": null,
-          "clean_text": ""
-        },
-        "author": null,
-        "producer": {
-          "text": "Peri",
-          "confidence": 0.8527556657791138,
-          "start": 1447,
-          "end": 1451,
-          "clean_text": "Peri"
-        },
-        "geography": {
-          "text": "Poland",
-          "confidence": 0.9999470710754395,
-          "start": 979,
-          "end": 985,
-          "clean_text": "Poland"
-        },
-        "publication_year": null,
-        "reference_year": {
-          "text": "2016",
-          "confidence": 0.6385188102722168,
-          "start": 1453,
-          "end": 1457
-        },
-        "reference_population": null,
-        "is_used": "False",
-        "usage_context": "background",
-        "validations": []
-      },
       {
         "dataset_name": {
           "text": "impact of Ukrainian refugees on Polish GDP",
@@ -5733,141 +5103,6 @@
         "usage_context": "background",
         "validations": []
       },
-      {
-        "dataset_name": {
-          "text": "Monitor Deloitte (2022)",
-          "confidence": 0.654185950756073,
-          "start": 2237,
-          "end": 2260,
-          "clean_text": "Monitor Deloitte (2022)"
-        },
-        "dataset_tag": "named",
-        "description": null,
-        "data_type": "estimates",
-        "acronym": {
-          "text": "",
-          "confidence": 0.0,
-          "start": 2237,
-          "end": 2260,
-          "clean_text": ""
-        },
-        "author": null,
-        "producer": {
-          "text": "",
-          "confidence": 0.0,
-          "start": 2237,
-          "end": 2260,
-          "clean_text": ""
-        },
-        "geography": {
-          "text": "",
-          "confidence": 0.0,
-          "start": 2237,
-          "end": 2260,
-          "clean_text": ""
-        },
-        "publication_year": null,
-        "reference_year": {
-          "text": "2022",
-          "confidence": 0.931653618812561,
-          "start": 2283,
-          "end": 2287
-        },
-        "reference_population": null,
-        "is_used": "False",
-        "usage_context": "background",
-        "validations": []
-      },
-      {
-        "dataset_name": {
-          "text": "Oxford Economics (2022)",
-          "confidence": 0.8788336515426636,
-          "start": 2265,
-          "end": 2288,
-          "clean_text": "Oxford Economics (2022)"
-        },
-        "dataset_tag": "named",
-        "description": null,
-        "data_type": "other",
-        "acronym": {
-          "text": "",
-          "confidence": 0.0,
-          "start": 2265,
-          "end": 2288,
-          "clean_text": ""
-        },
-        "author": null,
-        "producer": {
-          "text": "",
-          "confidence": 0.0,
-          "start": 2265,
-          "end": 2288,
-          "clean_text": ""
-        },
-        "geography": {
-          "text": "",
-          "confidence": 0.0,
-          "start": 2265,
-          "end": 2288,
-          "clean_text": ""
-        },
-        "publication_year": null,
-        "reference_year": {
-          "text": "2022",
-          "confidence": 0.8991004228591919,
-          "start": 2283,
-          "end": 2287
-        },
-        "reference_population": null,
-        "is_used": "True",
-        "usage_context": "supporting",
-        "validations": []
-      },
-      {
-        "dataset_name": {
-          "text": "Peri (2016, 2014)",
-          "confidence": 0.7804762721061707,
-          "start": 2741,
-          "end": 2758,
-          "clean_text": "Peri (2016, 2014)"
-        },
-        "dataset_tag": "named",
-        "description": null,
-        "data_type": "other",
-        "acronym": {
-          "text": "",
-          "confidence": 0.7748786807060242,
-          "start": null,
-          "end": null,
-          "clean_text": ""
-        },
-        "author": null,
-        "producer": {
-          "text": "",
-          "confidence": 0.0,
-          "start": 2741,
-          "end": 2758,
-          "clean_text": ""
-        },
-        "geography": {
-          "text": "",
-          "confidence": 0.0,
-          "start": 2741,
-          "end": 2758,
-          "clean_text": ""
-        },
-        "publication_year": null,
-        "reference_year": {
-          "text": "",
-          "confidence": 0.0,
-          "start": 2741,
-          "end": 2758
-        },
-        "reference_population": null,
-        "is_used": "False",
-        "usage_context": "background",
-        "validations": []
-      },
       {
         "dataset_name": {
           "text": "directed technical change",
@@ -6057,98 +5292,8 @@
     }
   },
   {
-    "input_text": "Analysis of the impact of refugees from Ukraine on the economy of Poland\n39\nAnalysis of the impact of refugees from Ukraine on the economy of Poland\n38\nProductivity shock rationale\nAs Ukrainian refugees entered the labour \nmarket, the economy adapted in line \nwith expectations based on scientific \nliterature, resulting in greater specialisation \nand higher productivity. In a simplistic \nsupply-demand framework akin to the \n“canonical model”, the influx of Ukrainian \nrefugees should have caused some Polish \nworkers to become unemployed or leave \nthe labour force, or real wages to fall. \nEven in the Deloitte D.Climate model, \nUkrainian refugees add 0.4 percentage \npoints to the unemployment rate and \nlower real wages by 1.35% in 2024 when \nit is not counterbalanced by an additional \npositive productivity shock. However, \nthis is not what we observe in empirical \ndata. First, Polish citizens employment \nrates have grown, and unemployment \nrates have fallen. Second, poviats in \nwhich the employment share of Ukrainian \nrefugees has grown by 1 percentage point, \nexperienced higher by 0.5 percentage \npoint Polish citizens employment rates, and \n0.3 percentage point lower unemployment \nrates. Third, there is no evidence of \nlowered wages; in fact, the limited available \ndata suggests that Ukrainian refugees \nmay have caused higher wage growth in \npoviats to which they have moved. These \nare common findings in the scientific \nliterature quoted in the previous section, \nthat as immigrants enter the labour \nmarket, native workers tend to specialise \nin higher-value, complementary tasks. \nThis is evident in Polish workers moving to \nmore attractive occupational groups. This \nconstitutes a positive productivity shock, \ncounterbalancing labour market pressures. \nFirst, the employment rate of Polish citizens \nhas been growing since the Ukrainian \nrefugee influx, while the unemployment \nrate has been falling. This is particularly \nevident among women, who make up the \nmajority of refugees. The employment \nrate for women aged 20-64 has grown \nconsistently from 68.2% in Q2 2021 to \n70.2% in Q2 2022, 71.7% in Q2 2023, and \n72.2% in Q2 2024 (see Chart 26). The \nunemployment rate for women aged \n20-64 has consistently fallen, from 3.5% in \nQ2 2021 to 3.0% in Q2 2022, 2.6% in both \nQ2 2023 and Q2 2024.\nSecond, panel regression analysis shows \nthat a larger influx of Ukrainian refugees \nat the poviat level was accompanied by \na larger increase in employment rates \nfor Polish citizens along with a larger \ndecrease in registered unemployment \nrates. The data sample for all regressions \nencompassed quarterly data for all \n380 poviats from Q1 2022 to Q2 2024. \nThe Ukrainian refugee share variable was \nconstructed as the share of Ukrainian \nrefugees among all employed, temporarily \nemployed, and self-employed persons \ninsured with ZUS in a given poviat at the \nend of the quarter. The employment rate \nof Polish citizens was constructed as the \nshare of Polish citizens who are employed, \ntemporarily employed, and self-employed \nand who are insured with ZUS, divided by \nthe population from GUS in a given poviat \nin a given quarter (estimated based on half-\nyear population data). The unemployment \nrate is the registered unemployment \nrate series from GUS. Fixed effects panel \nregressions show that an increase in the \nemployment share of Ukrainian refugees \nby 1 percentage point correlates with an \nincrease in Polish citizens' employment \nrates by 0.5% and a decrease in the \nunemployment rate by 0.3%, in the \npreferred specification with dummy \nvariables for all quarters. All results are \nstatistically significant at a 0.01 level.\nChart 26. Polish citizens employment rate  \nChart 27. Polish citizens unemployment rate\n2021-Q2\n2021-Q2\n2022-Q2\n2022-Q2\n2023-Q2\n2023-Q2\n2024-Q2\n2024-Q2\nSource: Deloitte own elaboration based \nof Eurostat data (Labour Force Survey).\nSource: Deloitte own elaboration based \nof Eurostat data (Labour Force Survey).\nEurostat survey data, 20-64 age group\nEurostat survey data, 20-64 age group\nMales\nMales\nFemales\nFemales\n83.8%\n72.2%\n2.6%\n2.6%\n83.6%\n71.7%\n2.6%\n3.0%\n3.5%\n2.5%\n3.6%\n2.3%\n83.2%\n70.2%\n82.0%\n68.2%\nChart 28. Effect of a 1 pp. change in employment share of Ukrainian refugees on\nNo time \ndummies\nNo time \ndummies\nSeasonal \ndummies\nSeasonal \ndummies\nQuarterly \ndummies\nQuarterly \ndummies\nSource: Deloitte own elaboration based on GUS and ZUS data. All continuous variables have been regressed in first differences to account for non-stationarity Polish \ncitizens employment rates and registered unemployment rates. For details see the Online Technical Appendix.\n-2.0\n-2.0\n-1.0\n-1.0\n0\n0\n1.0\n1.0\nPanel model of all 380 poviats quarterly data from Q1 2022 to Q2 2024. Results are statistically significant at a 0.01 level.\nPolish citizens employment rate\nUnemployment rate\n",
-    "datasets": [
-      {
-        "dataset_name": {
-          "text": "Deloitte D.Climate model",
-          "confidence": 0.39529725909233093,
-          "start": 595,
-          "end": 619,
-          "clean_text": "Deloitte D.Climate model"
-        },
-        "dataset_tag": "vague",
-        "description": null,
-        "data_type": "other",
-        "acronym": {
-          "text": "",
-          "confidence": 0.916789174079895,
-          "start": null,
-          "end": null,
-          "clean_text": ""
-        },
-        "author": null,
-        "producer": {
-          "text": "Deloitte D.",
-          "confidence": 0.7124313712120056,
-          "start": 595,
-          "end": 606,
-          "clean_text": "Deloitte D."
-        },
-        "geography": {
-          "text": "Ukraine",
-          "confidence": 0.9686239361763,
-          "start": 40,
-          "end": 47,
-          "clean_text": "Ukraine"
-        },
-        "publication_year": null,
-        "reference_year": {
-          "text": "",
-          "confidence": 0.0,
-          "start": 595,
-          "end": 619
-        },
-        "reference_population": null,
-        "is_used": "False",
-        "usage_context": "background",
-        "validations": []
-      },
-      {
-        "dataset_name": {
-          "text": "poviats",
-          "confidence": 0.30133524537086487,
-          "start": 958,
-          "end": 965,
-          "clean_text": "poviats"
-        },
-        "dataset_tag": "vague",
-        "description": null,
-        "data_type": "other",
-        "acronym": {
-          "text": "",
-          "confidence": 0.0,
-          "start": 958,
-          "end": 965,
-          "clean_text": ""
-        },
-        "author": null,
-        "producer": {
-          "text": "",
-          "confidence": 0.0,
-          "start": 958,
-          "end": 965,
-          "clean_text": ""
-        },
-        "geography": {
-          "text": "",
-          "confidence": 0.0,
-          "start": 958,
-          "end": 965,
-          "clean_text": ""
-        },
-        "publication_year": null,
-        "reference_year": {
-          "text": "",
-          "confidence": 0.0,
-          "start": 958,
-          "end": 965
-        },
-        "reference_population": null,
-        "is_used": "False",
-        "usage_context": "background",
-        "validations": []
-      },
+    "input_text": "Analysis of the impact of refugees from Ukraine on the economy of Poland\n39\nAnalysis of the impact of refugees from Ukraine on the economy of Poland\n38\nProductivity shock rationale\nAs Ukrainian refugees entered the labour \nmarket, the economy adapted in line \nwith expectations based on scientific \nliterature, resulting in greater specialisation \nand higher productivity. In a simplistic \nsupply-demand framework akin to the \n“canonical model”, the influx of Ukrainian \nrefugees should have caused some Polish \nworkers to become unemployed or leave \nthe labour force, or real wages to fall. \nEven in the Deloitte D.Climate model, \nUkrainian refugees add 0.4 percentage \npoints to the unemployment rate and \nlower real wages by 1.35% in 2024 when \nit is not counterbalanced by an additional \npositive productivity shock. However, \nthis is not what we observe in empirical \ndata. First, Polish citizens employment \nrates have grown, and unemployment \nrates have fallen. Second, poviats in \nwhich the employment share of Ukrainian \nrefugees has grown by 1 percentage point, \nexperienced higher by 0.5 percentage \npoint Polish citizens employment rates, and \n0.3 percentage point lower unemployment \nrates. Third, there is no evidence of \nlowered wages; in fact, the limited available \ndata suggests that Ukrainian refugees \nmay have caused higher wage growth in \npoviats to which they have moved. These \nare common findings in the scientific \nliterature quoted in the previous section, \nthat as immigrants enter the labour \nmarket, native workers tend to specialise \nin higher-value, complementary tasks. \nThis is evident in Polish workers moving to \nmore attractive occupational groups. This \nconstitutes a positive productivity shock, \ncounterbalancing labour market pressures. \nFirst, the employment rate of Polish citizens \nhas been growing since the Ukrainian \nrefugee influx, while the unemployment \nrate has been falling. This is particularly \nevident among women, who make up the \nmajority of refugees. The employment \nrate for women aged 20-64 has grown \nconsistently from 68.2% in Q2 2021 to \n70.2% in Q2 2022, 71.7% in Q2 2023, and \n72.2% in Q2 2024 (see Chart 26). The \nunemployment rate for women aged \n20-64 has consistently fallen, from 3.5% in \nQ2 2021 to 3.0% in Q2 2022, 2.6% in both \nQ2 2023 and Q2 2024.\nSecond, panel regression analysis shows \nthat a larger influx of Ukrainian refugees \nat the poviat level was accompanied by \na larger increase in employment rates \nfor Polish citizens along with a larger \ndecrease in registered unemployment \nrates. The data sample for all regressions \nencompassed quarterly data for all \n380 poviats from Q1 2022 to Q2 2024. \nThe Ukrainian refugee share variable was \nconstructed as the share of Ukrainian \nrefugees among all employed, temporarily \nemployed, and self-employed persons \ninsured with ZUS in a given poviat at the \nend of the quarter. The employment rate \nof Polish citizens was constructed as the \nshare of Polish citizens who are employed, \ntemporarily employed, and self-employed \nand who are insured with ZUS, divided by \nthe population from GUS in a given poviat \nin a given quarter (estimated based on half-\nyear population data). The unemployment \nrate is the registered unemployment \nrate series from GUS. Fixed effects panel \nregressions show that an increase in the \nemployment share of Ukrainian refugees \nby 1 percentage point correlates with an \nincrease in Polish citizens' employment \nrates by 0.5% and a decrease in the \nunemployment rate by 0.3%, in the \npreferred specification with dummy \nvariables for all quarters. All results are \nstatistically significant at a 0.01 level.\nChart 26. Polish citizens employment rate  \nChart 27. Polish citizens unemployment rate\n2021-Q2\n2021-Q2\n2022-Q2\n2022-Q2\n2023-Q2\n2023-Q2\n2024-Q2\n2024-Q2\nSource: Deloitte own elaboration based \nof Eurostat data (Labour Force Survey).\nSource: Deloitte own elaboration based \nof Eurostat data (Labour Force Survey).\nEurostat survey data, 20-64 age group\nEurostat survey data, 20-64 age group\nMales\nMales\nFemales\nFemales\n83.8%\n72.2%\n2.6%\n2.6%\n83.6%\n71.7%\n2.6%\n3.0%\n3.5%\n2.5%\n3.6%\n2.3%\n83.2%\n70.2%\n82.0%\n68.2%\nChart 28. Effect of a 1 pp. change in employment share of Ukrainian refugees on\nNo time \ndummies\nNo time \ndummies\nSeasonal \ndummies\nSeasonal \ndummies\nQuarterly \ndummies\nQuarterly \ndummies\nSource: Deloitte own elaboration based on GUS and ZUS data. All continuous variables have been regressed in first differences to account for non-stationarity Polish \ncitizens employment rates and registered unemployment rates. For details see the Online Technical Appendix.\n-2.0\n-2.0\n-1.0\n-1.0\n0\n0\n1.0\n1.0\nPanel model of all 380 poviats quarterly data from Q1 2022 to Q2 2024. Results are statistically significant at a 0.01 level.\nPolish citizens employment rate\nUnemployment rate\n",
+    "datasets": [
       {
         "dataset_name": {
           "text": "quarterly data for all 380 poviats",
@@ -6880,51 +6025,6 @@
         "usage_context": "primary",
         "validations": []
       },
-      {
-        "dataset_name": {
-          "text": "PLN 75",
-          "confidence": 0.5854560136795044,
-          "start": 4435,
-          "end": 4441,
-          "clean_text": "PLN 75"
-        },
-        "dataset_tag": "named",
-        "description": null,
-        "data_type": "other",
-        "acronym": {
-          "text": "",
-          "confidence": 0.0,
-          "start": 4435,
-          "end": 4441,
-          "clean_text": ""
-        },
-        "author": null,
-        "producer": {
-          "text": "",
-          "confidence": 0.0,
-          "start": 4435,
-          "end": 4441,
-          "clean_text": ""
-        },
-        "geography": {
-          "text": "",
-          "confidence": 0.0,
-          "start": 4435,
-          "end": 4441,
-          "clean_text": ""
-        },
-        "publication_year": null,
-        "reference_year": {
-          "text": "",
-          "confidence": 0.0,
-          "start": 4435,
-          "end": 4441
-        },
-        "reference_population": null,
-        "is_used": "True",
-        "usage_context": "supporting",
-        "validations": []
-      },
       {
         "dataset_name": {
           "text": "GUS data",
@@ -7431,51 +6531,6 @@
         "usage_context": "primary",
         "validations": []
       },
-      {
-        "dataset_name": {
-          "text": "UNHCR’s 2023 MSNA",
-          "confidence": 0.9005851149559021,
-          "start": 3296,
-          "end": 3313,
-          "clean_text": "UNHCR’s 2023 MSNA"
-        },
-        "dataset_tag": "named",
-        "description": null,
-        "data_type": "survey",
-        "acronym": {
-          "text": "",
-          "confidence": 0.5884755849838257,
-          "start": null,
-          "end": null,
-          "clean_text": ""
-        },
-        "author": null,
-        "producer": {
-          "text": "UNHCR",
-          "confidence": 0.8781561851501465,
-          "start": 3296,
-          "end": 3301,
-          "clean_text": "UNHCR"
-        },
-        "geography": {
-          "text": "",
-          "confidence": 0.0,
-          "start": 3296,
-          "end": 3313,
-          "clean_text": ""
-        },
-        "publication_year": null,
-        "reference_year": {
-          "text": "2023",
-          "confidence": 0.7625691890716553,
-          "start": 2809,
-          "end": 2813
-        },
-        "reference_population": null,
-        "is_used": "True",
-        "usage_context": "supporting",
-        "validations": []
-      },
       {
         "dataset_name": {
           "text": "2024 SEIS survey",
@@ -7892,51 +6947,6 @@
         "usage_context": "primary",
         "validations": []
       },
-      {
-        "dataset_name": {
-          "text": "PESEL/PESEL UKR",
-          "confidence": 0.5166351199150085,
-          "start": 464,
-          "end": 479,
-          "clean_text": "PESEL/PESEL UKR"
-        },
-        "dataset_tag": "named",
-        "description": null,
-        "data_type": "other",
-        "acronym": {
-          "text": "",
-          "confidence": 0.0,
-          "start": 464,
-          "end": 479,
-          "clean_text": ""
-        },
-        "author": null,
-        "producer": {
-          "text": "",
-          "confidence": 0.0,
-          "start": 464,
-          "end": 479,
-          "clean_text": ""
-        },
-        "geography": {
-          "text": "",
-          "confidence": 0.0,
-          "start": 464,
-          "end": 479,
-          "clean_text": ""
-        },
-        "publication_year": null,
-        "reference_year": {
-          "text": "",
-          "confidence": 0.0,
-          "start": 464,
-          "end": 479
-        },
-        "reference_population": null,
-        "is_used": "True",
-        "usage_context": "primary",
-        "validations": []
-      },
       {
         "dataset_name": {
           "text": "An ID number of every Polish citizen",
@@ -7982,51 +6992,6 @@
         "usage_context": "primary",
         "validations": []
       },
-      {
-        "dataset_name": {
-          "text": "PESEL UKR",
-          "confidence": 0.32711294293403625,
-          "start": 636,
-          "end": 645,
-          "clean_text": "PESEL UKR"
-        },
-        "dataset_tag": "named",
-        "description": null,
-        "data_type": "other",
-        "acronym": {
-          "text": "",
-          "confidence": 0.0,
-          "start": 636,
-          "end": 645,
-          "clean_text": ""
-        },
-        "author": null,
-        "producer": {
-          "text": "",
-          "confidence": 0.0,
-          "start": 636,
-          "end": 645,
-          "clean_text": ""
-        },
-        "geography": {
-          "text": "",
-          "confidence": 0.0,
-          "start": 636,
-          "end": 645,
-          "clean_text": ""
-        },
-        "publication_year": null,
-        "reference_year": {
-          "text": "",
-          "confidence": 0.0,
-          "start": 636,
-          "end": 645
-        },
-        "reference_population": null,
-        "is_used": "True",
-        "usage_context": "primary",
-        "validations": []
-      },
       {
         "dataset_name": {
           "text": "SEIS",
@@ -8218,51 +7183,6 @@
   {
     "input_text": "Analysis of the impact of refugees from Ukraine on the economy of Poland\n47\nAnalysis of the impact of refugees from Ukraine on the economy of Poland\n46\nAnalysis (SEIS 2024). UNHCR, October, https://data.unhcr.org/en/documents/\ndetails/115045\nUNHCR (2025b). High employment rates, but low wages: a poverty assessment \nof Ukrainian refugees in neighboring countries, Regional Refugee Response for \nthe Ukraine Situation, Regional Bureau for Europe, UNHCR.\nUNHCR (2025c). Ukraine Multi-year Strategy 2025 – 2027, UNHCR, November.\nUrban M. (2022). Refugees will lift economy's potential, but challenges remain, \nResearch Briefing | Poland. Oxford Economics, https://www.oxfordeconomics.\ncom/resource/refugees-in-poland-will-lift-economys-potential-but-challenges-\nremain/\nLiterature\nAcemoglu, D. (2002). Directed Technical Change, Review of Economic Studies \n69(4):781-80.\nAleksynska, M., & Tritah, A. (2015). The Heterogeneity Of Immigrants, Host \nCountries' Income And Productivity: A Channel Accounting Approach. Economic \nInquiry, 53(1), 150-172.\nAleksynska, M., & Tritah, A. (2013). Occupation–education mismatch of immigrant \nworkers in Europe: Context and policies, Economics of Education Review \n36:229-244.\nBabakova, O., Gomon, T., Naranovich, K., & Oliński, P. (2024). Ułatwienia w \nzatrudnianiu cudzoziemców spoza UE, Kierunek Polska. Rekomendacje dla \npolskiej polityki migracyjnej, Forum Obywatelskiego Rozwoju, July, ISBN \n978-83-63408-04-6.\nBeerli, A., & Peri, G. (2018). The labor market effects of opening the border: \nEvidence from Switzerland, NBER Working Paper No. 21319.\nBrell, C., Dustmann, C., & Preston, I. (2020). The labor market integration of \nrefugee migrants in high-income countries. Journal of Economic Perspectives, \n34(1), 94-121.\nBrücker, H., Glitz, A., Lerche, A., & Romiti, A. (2021). Occupational recognition and \nimmigrant labor market outcomes. Journal of Labor Economics,39(2)\nCassidy, H., & Dacass, T. (2021). Occupational licensing and immigrants. The \nJournal of Law and Economics, 64(1), 1-28.\nDeloitte (2024). Analysis of the impact of refugees from Ukraine on the economy \nof Poland, UNHCR Reports and Assessments, March, https://data.unhcr.org/en/\ndocuments/details/106993\nFoged, M., & Peri, G. (2016). Immigrants' effect on native workers: New analysis \non longitudinal data. American Economic Journal: Applied Economics, 8(2), 1-34.\nFoged, M., Hasager, L., & Peri, G. (2024). Comparing the effects of policies for \nthe labor market integration of refugees. Journal of Labor Economics, 42(S1), \nS335-S377.\nGromadzki, J., & Lewandowski, P. (2023). Refugees from Ukraine on the Polish \nlabour market. Social Insurance. Theory and Practice, 155(4), 29-40.\nGUS (2024a). Structure of wages and salaries by occupations for October 2022, \nStatistical Office in Bydgoszcz, https://stat.gov.pl/en/topics/labour-market/\nworking-employed-wages-and-salaries-cost-of-labour/structure-of-wages-and-\nsalaries-by-occupations-for-october-2022,4,8.html\nGUS (2024b). Education in the school year 2023/2024 (preliminary data), \nStatistical Office in Gdańsk, https://stat.gov.pl/en/topics/education/education/\neducation-in-the-school-year-20232024-preliminary-data,13,2.html\nHeller, B. H., & Mumma, K. S. (2023). Immigrant integration in the United States: \nthe role of adult English language training. American Economic Journal: Economic \nPolicy, 15(3), 407-437.\nJaumotte, M. F., Koloskova, K., & Saxena, M. S. C. (2016). Impact of migration on \nincome levels in advanced economies. International Monetary Fund.\nKleiner, M. M., & Krueger, A. B. (2013). Analyzing the Extent and Influence of \nOccupational Licensing on the Labor Market. Journal of Labor Economics, 31(2), \nS173–S202. https://doi.org/10.1086/669060\nLessem, R., & Sanders, C. (2020). Immigrant wage growth in the United States: \nThe role of occupational upgrading. International Economic Review, 61(2), \n941-972.\nLewandowski, P., Górny, A., Krząkała, M., & Palczyńska, M. (2025). The Role of Job \nTask Degradation in Shaping Return Intentions: Evidence from Ukrainian War \nRefugees in Poland. IBS working paper, 01/2025. https://ibs.org.pl/wp-content/\nuploads/2025/03/Job_task_degradation_return_intentions_Ukrainian_refugees_\nIBS_WP_202501.pdf\nLewis, E. (2013). Immigration and production technology, Annual Review of \nEconomics 5(1):165-191\nLewis, E., & Peri, G. (2015). “Immigration and the Economy of Cities and Regions”, \n[in] Duranton, G., Henderson, J.V., Strange, W., Handbook of Regional and Urban \nEconomics, Volume 5A, North-Holland, Elsevier.\nMeager N., & Speckesser, S. (2011). Wages, productivity and employment: \nA review of theory and international data, Brussels: European Employment \nObservatory.\nMonitor Deloitte (2022). Refugees from Ukraine in Poland: Challenges and \npotential for integration. October.\nNBP (2024). Sytuacja życiowa i ekonomiczna migrantów z Ukrainy w Polsce w \n2024 r. Raport z badania ankietowego, Departament Statystyki, Narodowy Bank \nPolski, Warszawa.\nPeri, G. (2012). The effect of immigration on productivity: Evidence from US \nstates. Review of Economics and Statistics, 94(1), 348-358.\nPeri, G. (2014). Do immigrant workers depress the wages of native workers? IZA \nWorld of Labor, 42.\nPeri, G. (2016). Immigrants, productivity, and labor markets. Journal of economic \nperspectives, 30(4), 3-30.\nPeri, G., & Sparber, C. (2009). Task specialization, immigration, and wages. \nAmerican Economic Journal: Applied Economics, 1(3), 135-169.\nPeterson, B. D., Pandya, S. S., & Leblang, D. (2014). Doctors with borders: \noccupational licensing as an implicit barrier to high skill migration. Public Choice, \n160(1-2), 45–63. doi:10.1007/s11127-014-0152-8\nSchmid, L. (2023). The impact of host language proficiency on migrants’ \nemployment outcomes. American Economic Review: Insights, 5(4), 477-493.\nSohst, R., Tirado, T., Salgado, L., & Slootjes, J. (2024). Exploring Refugees’ \nIntentions to Return to Ukraine: Data Insights and Policy Responses. \nInternational Organization for Migration and Migration Policy Institute Europe.\nStrzelecki, P., Growiec, J., & Wyszyński, R. (2022). The contribution of immigration \nfrom Ukraine to economic growth in Poland. Review of World Economics, 158(2), \n365-399.\nTokariuk, O. (2025). Ukraine’s fight for its people. Strategies for refugee and \ndiaspora engagement, Ukraine Forum, Chatham House, February. \nUNHCR (2023). Poland: Multi-Sector Needs Assessment — Results Overview \n(MSNA 2023), October, https://data.unhcr.org/fr/documents/details/104427\nUNHCR (2025a). Poland: Socio-Economic Insights Survey in Poland - Results \n",
     "datasets": [
-      {
-        "dataset_name": {
-          "text": "SEIS 2024",
-          "confidence": 0.9402666687965393,
-          "start": 162,
-          "end": 171,
-          "clean_text": "SEIS 2024"
-        },
-        "dataset_tag": "named",
-        "description": null,
-        "data_type": "other",
-        "acronym": {
-          "text": "",
-          "confidence": 0.9784660935401917,
-          "start": null,
-          "end": null,
-          "clean_text": ""
-        },
-        "author": null,
-        "producer": {
-          "text": "UNHCR",
-          "confidence": 0.9712179899215698,
-          "start": 452,
-          "end": 457,
-          "clean_text": "UNHCR"
-        },
-        "geography": {
-          "text": "Ukraine",
-          "confidence": 0.9796717762947083,
-          "start": 398,
-          "end": 405,
-          "clean_text": "Ukraine"
-        },
-        "publication_year": null,
-        "reference_year": {
-          "text": "2024",
-          "confidence": 0.9505027532577515,
-          "start": 167,
-          "end": 171
-        },
-        "reference_population": null,
-        "is_used": "True",
-        "usage_context": "primary",
-        "validations": []
-      },
       {
         "dataset_name": {
           "text": "Rekomendacje dla\npolskiej polityki migracyjnej",
@@ -8488,51 +7408,6 @@
         "usage_context": "background",
         "validations": []
       },
-      {
-        "dataset_name": {
-          "text": "Monitor Deloitte",
-          "confidence": 0.3540622591972351,
-          "start": 4668,
-          "end": 4684,
-          "clean_text": "Monitor Deloitte"
-        },
-        "dataset_tag": "named",
-        "description": null,
-        "data_type": "other",
-        "acronym": {
-          "text": "",
-          "confidence": 0.8403723835945129,
-          "start": null,
-          "end": null,
-          "clean_text": ""
-        },
-        "author": null,
-        "producer": {
-          "text": "",
-          "confidence": 0.0,
-          "start": 4668,
-          "end": 4684,
-          "clean_text": ""
-        },
-        "geography": {
-          "text": "Ukraine",
-          "confidence": 0.8352624773979187,
-          "start": 4707,
-          "end": 4714,
-          "clean_text": "Ukraine"
-        },
-        "publication_year": null,
-        "reference_year": {
-          "text": "",
-          "confidence": 0.0,
-          "start": 4668,
-          "end": 4684
-        },
-        "reference_population": null,
-        "is_used": "False",
-        "usage_context": "background",
-        "validations": []
-      },
       {
         "dataset_name": {
           "text": "Multi-Sector Needs Assessment — Results Overview",
@@ -8678,96 +7553,6 @@
         "is_used": "True",
         "usage_context": "primary",
         "validations": []
-      },
-      {
-        "dataset_name": {
-          "text": "Fortune Global 500®",
-          "confidence": 0.4832018315792084,
-          "start": 764,
-          "end": 783,
-          "clean_text": "Fortune Global 500®"
-        },
-        "dataset_tag": "named",
-        "description": null,
-        "data_type": "other",
-        "acronym": {
-          "text": "",
-          "confidence": 0.0,
-          "start": 764,
-          "end": 783,
-          "clean_text": ""
-        },
-        "author": null,
-        "producer": {
-          "text": "",
-          "confidence": 0.0,
-          "start": 764,
-          "end": 783,
-          "clean_text": ""
-        },
-        "geography": {
-          "text": "",
-          "confidence": 0.0,
-          "start": 764,
-          "end": 783,
-          "clean_text": ""
-        },
-        "publication_year": null,
-        "reference_year": {
-          "text": "",
-          "confidence": 0.0,
-          "start": 764,
-          "end": 783
-        },
-        "reference_population": null,
-        "is_used": "False",
-        "usage_context": "background",
-        "validations": []
-      },
-      {
-        "dataset_name": {
-          "text": "QR code",
-          "confidence": 0.33300963044166565,
-          "start": 2259,
-          "end": 2266,
-          "clean_text": "QR code"
-        },
-        "dataset_tag": "named",
-        "description": null,
-        "data_type": "other",
-        "acronym": {
-          "text": "",
-          "confidence": 0.960260272026062,
-          "start": null,
-          "end": null,
-          "clean_text": ""
-        },
-        "author": null,
-        "producer": {
-          "text": "",
-          "confidence": 0.0,
-          "start": 2259,
-          "end": 2266,
-          "clean_text": ""
-        },
-        "geography": {
-          "text": "",
-          "confidence": 0.0,
-          "start": 2259,
-          "end": 2266,
-          "clean_text": ""
-        },
-        "publication_year": null,
-        "reference_year": {
-          "text": "",
-          "confidence": 0.0,
-          "start": 2259,
-          "end": 2266
-        },
-        "reference_population": null,
-        "is_used": "False",
-        "usage_context": "background",
-        "validations": []
       }
     ],
     "document": {