Upload 13 files

Initial MultiPII-X v1.0 dataset upload

.gitattributes

@@ -58,3 +58,5 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
 # Video files - compressed
 *.mp4 filter=lfs diff=lfs merge=lfs -text
 *.webm filter=lfs diff=lfs merge=lfs -text
+gliner/gliner_train.jsonl filter=lfs diff=lfs merge=lfs -text
+gliner25/gliner25_train.jsonl filter=lfs diff=lfs merge=lfs -text

CITATION.cff

@@ -0,0 +1,242 @@
+cff-version: 1.2.0
+
+message: >-
+  If you use MultiPII-X, please cite this dataset. MultiPII-X builds on and
+  substantially transforms several existing datasets and corpora; the original
+  source resources are credited in the references below.
+
+title: "MultiPII-X: A Fine-Grained Multi-Domain PII Dataset"
+
+type: dataset
+
+version: "1.0.0"
+
+date-released: "2026-10-03"
+
+authors:
+  - given-names: Nenad
+    family-names: Jakovchevski
+    affiliation: "Jožef Stefan Institute, Department of Knowledge Technologies"
+
+abstract: >-
+  MultiPII-X is an extensible fine-grained dataset for personally identifiable
+  information (PII), sensitive-entity recognition, and text anonymization.
+  Version 1.0 contains Slovene data spanning multiple domains and source
+  collections. The release contains 15,180 annotated texts with 89,812 entity
+  annotations across 56 fine-grained entity types. Source material was
+  translated, adapted, sampled, reannotated, or otherwise transformed under
+  a unified annotation ontology. The canonical release preserves original text
+  together with character-level entity spans and also provides derived
+  representations for GLiNER and GLiNER2/GLiNER2.5.
+
+keywords:
+  - personally identifiable information
+  - PII
+  - anonymization
+  - named entity recognition
+  - NER
+  - privacy
+  - sensitive information
+  - Slovene
+  - Slovenian
+  - low-resource language
+  - multi-domain NLP
+  - GLiNER
+
+references:
+
+  # --------------------------------------------------------------------------
+  # TANAOS
+  # --------------------------------------------------------------------------
+
+  - type: data
+    title: "Tanaos Text Anonymizer Training Dataset"
+    authors:
+      - name: "Tanaos"
+    database: "Hugging Face Hub"
+    entry: "tanaos/synthetic-text-anonymizer-dataset-v1"
+    license: MIT
+    scope: >-
+      Source dataset for MultiPII-X. Selected synthetic English PII examples
+      were adapted into Slovene and transformed to the MultiPII-X annotation
+      ontology.
+
+  # --------------------------------------------------------------------------
+  # KURKOWSKI
+  # --------------------------------------------------------------------------
+
+  - type: data
+    title: "Contextual Text Anonymizer Dataset"
+    authors:
+      - given-names: Michał
+        family-names: Kurkowski
+    year: 2025
+    database: "Hugging Face Hub"
+    entry: "kurkowski/synthetic-contextual-anonymizer-dataset"
+    license: MIT
+    scope: >-
+      Source dataset for MultiPII-X. Contextual synthetic PII examples were
+      adapted into Slovene and mapped/reannotated under the MultiPII-X
+      fine-grained ontology.
+
+  # --------------------------------------------------------------------------
+  # GIGAFIDA 2.0
+  # --------------------------------------------------------------------------
+
+  - type: data
+    title: "Corpus of Written Standard Slovene Gigafida 2.0"
+    authors:
+      - given-names: Simon
+        family-names: Krek
+      - given-names: Tomaž
+        family-names: Erjavec
+      - given-names: Andraž
+        family-names: Repar
+      - given-names: Jaka
+        family-names: Čibej
+      - given-names: Špela
+        family-names: Arhar Holdt
+      - given-names: Polona
+        family-names: Gantar
+      - given-names: Iztok
+        family-names: Kosem
+      - given-names: Marko
+        family-names: Robnik-Šikonja
+      - given-names: Nikola
+        family-names: Ljubešić
+      - given-names: Kaja
+        family-names: Dobrovoljc
+      - given-names: Cyprian
+        family-names: Laskowski
+      - given-names: Miha
+        family-names: Grčar
+      - given-names: Peter
+        family-names: Holozan
+      - given-names: Simon
+        family-names: Šuster
+      - given-names: Vojko
+        family-names: Gorjanc
+      - given-names: Marko
+        family-names: Stabej
+      - given-names: Nataša
+        family-names: Logar
+    date-released: "2019-06-13"
+    database: "CLARIN.SI"
+    entry: "hdl:11356/1320"
+    scope: >-
+      Source corpus for naturally occurring Slovene text in MultiPII-X.
+      Selected passages were independently annotated under the PII ontology;
+      the original Gigafida linguistic annotations are not treated as
+      MultiPII-X PII annotations.
+
+  # --------------------------------------------------------------------------
+  # TELECOM CONVERSATION CORPUS
+  # --------------------------------------------------------------------------
+
+  - type: data
+    title: "Telecom 200k Dataset"
+    authors:
+      - name: "eisenzopf"
+    database: "Hugging Face Hub"
+    entry: "eisenzopf/telecom-conversation-corpus"
+    license: MIT
+    scope: >-
+      Source corpus for telecom-domain conversational material. Selected
+      synthetic conversations were adapted into Slovene and annotated using
+      the MultiPII-X ontology.
+
+  # --------------------------------------------------------------------------
+  # E3-JSI SYNTHETIC MULTI-PII DATASET
+  # --------------------------------------------------------------------------
+
+  - type: data
+    title: "Synthetic Multi-Domain PII NER Dataset"
+    authors:
+      - name: "Department for Artificial Intelligence, Jožef Stefan Institute (E3-JSI)"
+    year: 2024
+    database: "Hugging Face Hub"
+    entry: "E3-JSI/synthetic-multi-pii-ner-v1"
+    doi: "10.57967/hf/3211"
+    license: MIT
+    scope: >-
+      Source dataset for MultiPII-X. The Slovene subset was selected and
+      reannotated using the expanded MultiPII-X entity ontology.
+
+  # --------------------------------------------------------------------------
+  # POVEJMO-VEMO-MED
+  # --------------------------------------------------------------------------
+
+  - type: data
+    title: "Dataset of Slovene Medical Texts PoVeJMo-VeMo-Med 1.0"
+    authors:
+      - given-names: Miha
+        family-names: Malenšek
+      - given-names: Marko
+        family-names: Bajec
+    date-released: "2024-09-25"
+    publisher:
+      name: "Faculty of Computer and Information Science, University of Ljubljana"
+    database: "CLARIN.SI"
+    entry: "hdl:11356/1983"
+    license: CC-BY-4.0
+    scope: >-
+      Source corpus for the medical domain. MultiPII-X uses sampled texts from
+      the deduplicated Slovene medical corpus, which were newly annotated with
+      the MultiPII-X PII, identifier, measurement, and health-related ontology.
+
+  # --------------------------------------------------------------------------
+  # COLESLAW
+  # --------------------------------------------------------------------------
+
+  - type: data
+    title: "Collection of Slovenian Legal Texts COLESLAW 1.0"
+    authors:
+      - given-names: Miha
+        family-names: Malenšek
+      - given-names: Slavko
+        family-names: Žitnik
+      - given-names: Aleš
+        family-names: Završnik
+      - given-names: Saša
+        family-names: Krajnc
+      - given-names: Primož
+        family-names: Križnar
+      - given-names: Marko
+        family-names: Bajec
+    date-released: "2024-12-31"
+    publisher:
+      name: "Faculty of Computer and Information Science, University of Ljubljana"
+    database: "CLARIN.SI"
+    entry: "hdl:11356/2095"
+    license: CC-BY-4.0
+    scope: >-
+      Source corpus for the legal domain. Selected Slovene legal passages were
+      newly annotated using the MultiPII-X fine-grained PII and sensitive-entity
+      ontology.
+
+  # --------------------------------------------------------------------------
+  # PRIOR SLOVENE PII BENCHMARK / ANNOTATION METHODOLOGY
+  # --------------------------------------------------------------------------
+
+  - type: conference-paper
+    title: >-
+      Privacy-Preserving NLP for a Low-Resource Language: Benchmarking
+      LLM-Based Personally Identifiable Information Anonymization in Slovene
+    authors:
+      - given-names: Nenad
+        family-names: Jakovchevski
+      - given-names: Matej
+        family-names: Martinc
+    year: 2026
+    conference:
+      name: "17th International Conference of the CLEF Association (CLEF 2026)"
+    collection-title: >-
+      Experimental IR Meets Multilinguality, Multimodality, and Interaction
+    start: 55
+    end: 68
+    publisher:
+      name: "Springer"
+    scope: >-
+      Prior Slovene PII benchmark and annotation methodology on which part of
+      the annotation conventions and earlier training-resource construction
+      underlying MultiPII-X are based.

README.md

@@ -1,3 +1,355 @@
 ---
-license: cc-by-4.0
+pretty_name: MultiPII-X
+language:
+- sl
+task_categories:
+- token-classification
+size_categories:
+- 10K<n<100K
+source_datasets:
+- extended
+tags:
+- text
+- pii
+- privacy
+- anonymization
+- named-entity-recognition
+- fine-grained-ner
+- slovene
+- multi-domain
+- medical
+- legal
+- telecom
+- news
+- finance
+- banking
+- gliner
+- gliner2
+configs:
+- config_name: canonical
+  default: true
+  data_files:
+  - split: train
+    path: canonical/slovene_pii_train.parquet
+- config_name: gliner
+  data_files:
+  - split: train
+    path: gliner/gliner_train.parquet
+- config_name: gliner25
+  data_files:
+  - split: train
+    path: gliner25/gliner25_train.parquet
 ---
+
+# MultiPII-X
+
+## Description
+
+**MultiPII-X** is an extensible fine-grained dataset for personally identifiable information (PII), sensitive-entity recognition, and text anonymization.
+
+Version **1.0** contains Slovene (`sl`) data collected and adapted from multiple source corpora and domains under one unified annotation scheme.
+
+The release contains:
+
+- **15,180** text records
+- **89,812** entity annotations
+- **56** fine-grained entity types
+- **8** normalized source collections
+- general, news, telecom, legal, medical, healthcare, finance, and banking domains
+
+The dataset is intended for training and research on fine-grained PII detection, anonymization, and named entity recognition.
+
+MultiPII-X does **not** prescribe an official train/validation/test partition. The complete release is exposed as one `train` split so users can construct splits appropriate to their experiments.
+
+## Dataset representations
+
+MultiPII-X provides three configurations.
+
+### `canonical`
+
+The canonical representation is the authoritative version of the dataset.
+
+```python
+from datasets import load_dataset
+
+dataset = load_dataset(
+    "Nan0NJ/MultiPII-X",
+    "canonical",
+    split="train",
+)
+```
+
+Each record contains:
+
+- `text` — original text
+- `tagged` — text with annotated spans replaced by `[ENTITY_TYPE]`
+- `source` — normalized provenance/source identifier
+- `domain` — text domain
+- `language` — ISO language code (`sl`)
+- `entities` — character-level entity annotations
+- `gliner_tokenized_text` — derived word tokens for classic GLiNER
+- `gliner_entities` — derived token-level entities where exact representation is possible
+
+Each canonical entity contains:
+
+```json
+{
+  "start": 0,
+  "end": 5,
+  "text": "Janez",
+  "entity_type": "NAME_GIVEN",
+  "entity_group": "PERSON"
+}
+```
+
+Character offsets are zero-based and use an exclusive end offset:
+
+```python
+text[start:end] == entity["text"]
+```
+
+The `entities` field is the source of truth for annotation boundaries.
+
+### `gliner`
+
+A convenience representation for classic [GLiNER](https://github.com/urchade/GLiNER).
+
+```python
+dataset = load_dataset(
+    "Nan0NJ/MultiPII-X",
+    "gliner",
+    split="train",
+)
+```
+
+The representation contains:
+
+- `tokenized_text`
+- `ner`
+
+Entity spans use zero-based token indices with an inclusive end index.
+
+The classic GLiNER export contains **14,631** complete examples. The remaining **549** canonical rows contain character-level entity boundaries that cannot be represented exactly by the classic word-token representation without modifying the annotation. They remain available in `canonical` and `gliner25`.
+
+The native JSONL representation is also provided in:
+
+```text
+gliner/gliner_train.jsonl
+```
+
+### `gliner25`
+
+A convenience representation for [GLiNER2 / GLiNER2.5](https://github.com/fastino-ai/GLiNER2).
+
+```python
+dataset = load_dataset(
+    "Nan0NJ/MultiPII-X",
+    "gliner25",
+    split="train",
+)
+```
+
+All **15,180** records are available in this representation.
+
+The native JSONL representation is also provided in:
+
+```text
+gliner25/gliner25_train.jsonl
+```
+
+## Dataset sources
+
+MultiPII-X is a derived and reannotated resource. The original source datasets remain separate works and should be cited according to their respective terms.
+
+The normalized `source` field contains the following values:
+
+- `tanaos` — **7,612** records
+- `gigafida` — **2,125** records
+- `eeisenzopf` — **1,494** records
+- `kurkowski` — **1,026** records
+- `COLESLAW` — **1,000** records
+- `PoVeJMo-VeMo-Med` — **1,000** records
+- `telekom_synthetic` — **500** records
+- `e3-jsi` — **423** records
+
+### Upstream resources
+
+- [Tanaos Text Anonymizer Training Dataset](https://huggingface.co/datasets/tanaos/synthetic-text-anonymizer-dataset-v1)
+- [Contextual Text Anonymizer Dataset](https://huggingface.co/datasets/kurkowski/synthetic-contextual-anonymizer-dataset)
+- [Corpus of Written Standard Slovene Gigafida 2.0](https://www.clarin.si/repository/xmlui/handle/11356/1320)
+- [Telecom Conversation Corpus](https://huggingface.co/datasets/eisenzopf/telecom-conversation-corpus)
+- [E3-JSI Synthetic Multi-Domain PII NER Dataset](https://huggingface.co/datasets/E3-JSI/synthetic-multi-pii-ner-v1)
+- [PoVeJMo-VeMo-Med 1.0](https://www.clarin.si/repository/xmlui/handle/11356/1983)
+- [COLESLAW 1.0](https://www.clarin.si/repository/xmlui/handle/11356/2095)
+
+`telekom_synthetic` is additional telecom-oriented synthetic material included in the MultiPII-X construction process rather than a separately published upstream dataset.
+
+Full citation metadata for upstream resources is provided in `CITATION.cff`.
+
+## Domains
+
+The current Slovene release contains the following domains:
+
+- `general` — 8,742 records
+- `news` — 2,125 records
+- `telecom` — 1,994 records
+- `legal` — 1,065 records
+- `medical` — 1,000 records
+- `banking` — 94 records
+- `finance` — 82 records
+- `healthcare` — 78 records
+
+The `e3-jsi` subset retains its original domain assignments.
+
+## Entity types
+
+MultiPII-X v1.0 defines **56 fine-grained entity types**.
+
+The fine-grained `entity_type` is the model target. `entity_group` is additional metadata and should not replace the fine-grained label during training.
+
+### PERSON
+
+`NAME_GIVEN`, `NAME_FAMILY`, `NAME_ALIAS`
+
+### TEMPORAL
+
+`DATE_OF_BIRTH`, `DATE`, `AGE`, `PASSPORT_EXPIRATION`, `CREDIT_CARD_EXPIRATION`
+
+### CONTACT
+
+`EMAIL_ADDRESS`, `PHONE_NUMBER`
+
+### LOCATION
+
+`LOCATION_ADDRESS`, `LOCATION_CITY`, `LOCATION_ZIP`, `LOCATION_OTHER`, `LOCATION_COORDINATES`
+
+### FINANCIAL
+
+`BANK_ACCOUNT`, `CREDIT_CARD`
+
+### IDENTIFIER
+
+`SSN`, `PASSPORT_NUMBER`, `INTERNAL_ID`, `IMEI`, `IMSI`, `RIO`, `HEALTH_INSURANCE_ID`, `DRIVER_LICENSE_NUMBER`, `STUDENT_ID`, `INSURANCE_ID`, `REGISTRATION_NUMBER`, `VEHICLE_REGISTRATION_NUMBER`, `SERIAL_NUMBER`, `RESERVATION_NUMBER`, `TRANSACTION_ID`, `VISA_NUMBER`, `BIRTH_CERTIFICATE_NUMBER`, `TRAIN_TICKET_NUMBER`, `IDENTITY_DOCUMENT_NUMBER`
+
+### DIGITAL
+
+`USERNAME`, `URL`, `IP_OR_MAC`
+
+### CONTEXTUAL
+
+`ORGANIZATION`, `PRODUCT`
+
+### SECRET
+
+`TAX_ID`, `PIN`, `PUK`, `CVV`, `PASSWORD`, `SIM_CODE`, `ACTIVATION_CODE`, `VERIFICATION_CODE`, `AUTH_TOKEN`
+
+### MISC
+
+`CURRENCY_AMOUNT`, `MEASUREMENT`, `RATIO`
+
+### HEALTH
+
+`MEDICAL_CONDITION`, `MEDICATION`, `BLOOD_TYPE`
+
+Detailed label definitions are available in `label_registry.json`.
+
+## Annotation conventions
+
+MultiPII-X uses exact span annotations under a unified fine-grained ontology.
+
+Important conventions include:
+
+- **Full entity spans.** The complete intended entity is annotated.
+- **Slovene suffix absorption.** Inflectional endings attached to protected entities are included in the entity span.
+- **Honorific exclusion.** Titles such as `gospod`, `gospa`, `dr.`, and `prof.` remain outside personal-name annotations.
+- **No clock-time tagging.** Clock times are not annotated as `DATE`.
+- **Fine-grained secret labels.** Tax IDs, PINs, PUKs, CVVs, passwords, SIM codes, activation codes, verification codes, and authentication tokens use dedicated labels.
+- **Fine-grained numeric labels.** Monetary amounts, measurements, and ratios use `CURRENCY_AMOUNT`, `MEASUREMENT`, and `RATIO`.
+
+Raw source text is preserved in `text`. Character-level entities are authoritative.
+
+## Construction
+
+The source material was translated, adapted, sampled, reannotated, or otherwise transformed as appropriate for each upstream resource and normalized into the common MultiPII-X ontology.
+
+The final release contains all **15,180** prepared records and all **89,812** final annotations.
+
+The construction pipeline validated:
+
+- entity labels against the frozen ontology
+- ordered and non-overlapping character spans
+- exact `text[start:end]` recovery
+- deterministic reconstruction of `tagged`
+- source, domain, and language normalization
+- derived classic GLiNER token spans
+- GLiNER2/2.5 representation compatibility
+
+Rare labels and zero-entity records are intentionally retained.
+
+The release contains **634** zero-entity records.
+
+## Intended use
+
+MultiPII-X is intended for:
+
+- fine-grained PII and sensitive-entity recognition
+- text anonymization and pseudonymization research
+- NER model training and fine-tuning
+- cross-domain PII detection
+- evaluation of domain adaptation and label generalization
+- GLiNER and GLiNER2/2.5 fine-tuning
+- research on privacy-preserving NLP for Slovene
+
+The canonical representation is model-agnostic and can be converted for architectures other than GLiNER.
+
+## Limitations
+
+- Version 1.0 contains Slovene only.
+- Entity frequencies are intentionally imbalanced.
+- Some fine-grained labels contain very few examples.
+- Several source collections are synthetic or translated and should not be treated as representative samples of naturally occurring Slovene.
+- Domains differ substantially in style, length, and annotation density.
+- The model-specific exports inherit representation-specific limitations from their target model formats.
+- MultiPII-X is a research resource and does not itself establish legal or regulatory compliance for deployed anonymization systems.
+
+## License
+
+MultiPII-X incorporates transformed material from several upstream resources.
+
+A repository-level license should only be added after confirming that it is compatible with the redistribution terms of every included source. The original resources retain their respective attribution and licensing requirements.
+
+See the upstream resources above and `CITATION.cff` for provenance.
+
+## Citation
+
+If you use MultiPII-X, please cite the dataset version used.
+
+Citation metadata is provided in `CITATION.cff`.
+
+Until a permanent DOI is assigned, a provisional citation is:
+
+```bibtex
+@dataset{jakovchevski2026multipiix,
+  author    = {Nenad Jakovchevski},
+  title     = {MultiPII-X},
+  year      = {2026},
+  version   = {1.0.0},
+  publisher = {Hugging Face},
+  url       = {https://huggingface.co/datasets/Nan0NJ/MultiPII-X}
+}
+```
+
+## Author
+
+**Nenad Jakovchevski**  
+Department of Knowledge Technologies  
+Jožef Stefan Institute  
+Ljubljana, Slovenia
+
+## Acknowledgements
+
+MultiPII-X builds on multiple existing datasets and corpora. Their creators and maintainers are acknowledged in `CITATION.cff`.
+
+Part of the annotation methodology builds on prior work on fine-grained PII anonymization in Slovene:
+
+*Nenad Jakovchevski and Matej Martinc. Privacy-Preserving NLP for a Low-Resource Language: Benchmarking LLM-Based Personally Identifiable Information Anonymization in Slovene. CLEF 2026.*

canonical/slovene_pii_train.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1a8dfd98e2676ae59e84d6b571272775c0d8736806751f7fc28c3c0180606eb6
+size 8583317

dataset_statistics.json

@@ -0,0 +1,266 @@
+{
+  "dataset_name": "Slovene PII 56 (working title)",
+  "language": "sl",
+  "input_rows": 15180,
+  "canonical_rows": 15180,
+  "quarantine_rows": 0,
+  "annotations": 89812,
+  "registry_label_count": 56,
+  "observed_labels": 56,
+  "zero_entity_rows": 634,
+  "classic_gliner_rows": 14631,
+  "classic_only_exclusions": 549,
+  "gliner25_rows": 15180,
+  "rows_by_source": {
+    "COLESLAW": 1000,
+    "PoVeJMo-VeMo-Med": 1000,
+    "e3-jsi": 423,
+    "eeisenzopf": 1494,
+    "gigafida": 2125,
+    "kurkowski": 1026,
+    "tanaos": 7612,
+    "telekom_synthetic": 500
+  },
+  "rows_by_domain": {
+    "banking": 94,
+    "finance": 82,
+    "general": 8742,
+    "healthcare": 78,
+    "legal": 1065,
+    "medical": 1000,
+    "news": 2125,
+    "telecom": 1994
+  },
+  "entities_by_type": {
+    "NAME_GIVEN": 17667,
+    "NAME_FAMILY": 6117,
+    "NAME_ALIAS": 6,
+    "DATE_OF_BIRTH": 518,
+    "DATE": 8340,
+    "AGE": 37,
+    "PASSPORT_EXPIRATION": 6,
+    "CREDIT_CARD_EXPIRATION": 25,
+    "EMAIL_ADDRESS": 1280,
+    "PHONE_NUMBER": 3376,
+    "LOCATION_ADDRESS": 3003,
+    "LOCATION_CITY": 7620,
+    "LOCATION_ZIP": 1266,
+    "LOCATION_OTHER": 5327,
+    "LOCATION_COORDINATES": 47,
+    "BANK_ACCOUNT": 340,
+    "CREDIT_CARD": 104,
+    "SSN": 391,
+    "PASSPORT_NUMBER": 20,
+    "INTERNAL_ID": 1811,
+    "IMEI": 125,
+    "IMSI": 120,
+    "RIO": 95,
+    "HEALTH_INSURANCE_ID": 32,
+    "DRIVER_LICENSE_NUMBER": 8,
+    "STUDENT_ID": 12,
+    "INSURANCE_ID": 24,
+    "REGISTRATION_NUMBER": 232,
+    "VEHICLE_REGISTRATION_NUMBER": 17,
+    "SERIAL_NUMBER": 2,
+    "RESERVATION_NUMBER": 8,
+    "TRANSACTION_ID": 10,
+    "VISA_NUMBER": 7,
+    "BIRTH_CERTIFICATE_NUMBER": 5,
+    "TRAIN_TICKET_NUMBER": 6,
+    "IDENTITY_DOCUMENT_NUMBER": 16,
+    "USERNAME": 170,
+    "URL": 173,
+    "IP_OR_MAC": 167,
+    "ORGANIZATION": 8840,
+    "PRODUCT": 2196,
+    "TAX_ID": 252,
+    "PIN": 466,
+    "PUK": 3,
+    "CVV": 8,
+    "PASSWORD": 85,
+    "SIM_CODE": 1,
+    "ACTIVATION_CODE": 9,
+    "VERIFICATION_CODE": 7,
+    "AUTH_TOKEN": 1,
+    "CURRENCY_AMOUNT": 1230,
+    "MEASUREMENT": 5909,
+    "RATIO": 1235,
+    "MEDICAL_CONDITION": 4089,
+    "MEDICATION": 6943,
+    "BLOOD_TYPE": 8
+  },
+  "entities_by_group": {
+    "CONTACT": 4656,
+    "CONTEXTUAL": 11036,
+    "DIGITAL": 510,
+    "FINANCIAL": 444,
+    "HEALTH": 11040,
+    "IDENTIFIER": 2941,
+    "LOCATION": 17263,
+    "MISC": 8374,
+    "PERSON": 23790,
+    "SECRET": 832,
+    "TEMPORAL": 8926
+  },
+  "manual_adjudication": {
+    "manually_adjudicated_rows": 221,
+    "manual_recovered_entities": 3433,
+    "manual_historical_tagged_repairs": 28,
+    "previously_valid_tagged_whitespace_repairs": 524,
+    "total_historical_tagged_repairs": 552,
+    "manual_confidence_counts": {
+      "HIGH": 211,
+      "LOW": 1,
+      "MEDIUM": 9
+    }
+  },
+  "occurrence_ambiguity": {
+    "repeated_same_label_surface_rows": 3544,
+    "repeated_mentions_retained": 18743,
+    "rows_with_extra_exact_surface_occurrences": 1424,
+    "label_surface_pairs_with_extra_exact_occurrences": 2202,
+    "official_word_splitter_audit": {
+      "available": true,
+      "package_version": "2.0.0",
+      "class": "gliner2.processing.word_splitter.WhitespaceTokenSplitter",
+      "source_sha256": "8c533d4defec1cc469578bdb87244d57a82684ecff57431d0556a54e9a6eefc8",
+      "lower": true
+    },
+    "rows_with_extra_official_token_matches": 1017,
+    "label_surface_pairs_with_extra_official_token_matches": 1966,
+    "label_surface_pairs_not_exactly_recoverable_by_official_token_matches": 1098,
+    "behavior": "Mention lists retain every annotated occurrence. Official GLiNER2 resolves each surface to all matching word-token occurrences using lower=True. Mention JSONL cannot preserve occurrence-specific char offsets; Parquet entities remains authoritative.",
+    "zero_entity_behavior": "All frozen labels with empty mention lists form a valid negative entity task.",
+    "reference": "https://github.com/fastino-ai/GLiNER2/blob/main/gliner2/processor.py"
+  },
+  "licence_status": "Requires final confirmation; no uniform release licence established",
+  "authoritative_fields": [
+    "text",
+    "tagged",
+    "source",
+    "domain",
+    "language",
+    "entities"
+  ],
+  "derived_master_fields": [
+    "gliner_tokenized_text",
+    "gliner_entities"
+  ],
+  "offset_convention": "zero-based Python Unicode code points, half-open [start, end)",
+  "deduplicated": false,
+  "dataset_splits": {
+    "train": 15180
+  },
+  "source_files": [
+    {
+      "piece": 1,
+      "source_file": "ready_for_conversion_1.csv",
+      "sha256": "42d04b1f71bb82891793e45aee22adc7569131080a80bb9f903c9adf4f8d3726",
+      "rows": 10937,
+      "converted": 10937,
+      "quarantined": 0,
+      "columns": [
+        "_row_id",
+        "source",
+        "slo_raw",
+        "SLO_TAGGED",
+        "_secret_count",
+        "_misc_count",
+        "_needs_secret_refinement",
+        "_needs_misc_refinement",
+        "_needs_gemini_review",
+        "_split_reason"
+      ],
+      "annotation_column": "SLO_TAGGED"
+    },
+    {
+      "piece": 2,
+      "source_file": "ready_for_conversion_2.csv",
+      "sha256": "26e5fd6fffe102f700d005baffe72d42eaa140b2bbe5cab8224ba02aea165bea",
+      "rows": 1448,
+      "converted": 1448,
+      "quarantined": 0,
+      "columns": [
+        "_row_id",
+        "source",
+        "slo_raw",
+        "SLO_TAGGED_ORIGINAL",
+        "SLO_TAGGED_REFINED",
+        "target_count",
+        "decisions"
+      ],
+      "annotation_column": "SLO_TAGGED_REFINED"
+    },
+    {
+      "piece": 3,
+      "source_file": "ready_for_conversion_3.csv",
+      "sha256": "1409ffe39d0926eb574d9113041790eca8dd37936c38b8195ad6c41b1f2afa38",
+      "rows": 372,
+      "converted": 372,
+      "quarantined": 0,
+      "columns": [
+        "_row_id",
+        "source",
+        "slo_raw",
+        "SLO_TAGGED"
+      ],
+      "annotation_column": "SLO_TAGGED"
+    },
+    {
+      "piece": 4,
+      "source_file": "ready_for_conversion_4.csv",
+      "sha256": "dfb33d56fa6dcecef3faaf930522aa0aac21ff61340bb6c851def301086b66a5",
+      "rows": 423,
+      "converted": 423,
+      "quarantined": 0,
+      "columns": [
+        "_row_id",
+        "source",
+        "language",
+        "domain",
+        "slo_raw",
+        "SLO_TAGGED"
+      ],
+      "annotation_column": "SLO_TAGGED"
+    },
+    {
+      "piece": 5,
+      "source_file": "ready_for_conversion_5.csv",
+      "sha256": "e052f17263ea482070deed12ded85b17215bacc960a4121bddf2cfc244653ba3",
+      "rows": 1000,
+      "converted": 1000,
+      "quarantined": 0,
+      "columns": [
+        "original_id",
+        "source",
+        "source_file",
+        "language",
+        "domain",
+        "subdomain",
+        "slo_raw",
+        "SLO_TAGGED"
+      ],
+      "annotation_column": "SLO_TAGGED"
+    },
+    {
+      "piece": 6,
+      "source_file": "ready_for_conversion_6.csv",
+      "sha256": "7840b5deb1611fbccd60a49e194964544a817b5013de3c7f547b099bf018b5b5",
+      "rows": 1000,
+      "converted": 1000,
+      "quarantined": 0,
+      "columns": [
+        "original_id",
+        "source",
+        "source_file",
+        "language",
+        "domain",
+        "subdomain",
+        "title",
+        "slo_raw",
+        "SLO_TAGGED"
+      ],
+      "annotation_column": "SLO_TAGGED"
+    }
+  ]
+}

entities_by_group.csv

@@ -0,0 +1,12 @@
+entity_group,entities
+CONTACT,4656
+CONTEXTUAL,11036
+DIGITAL,510
+FINANCIAL,444
+HEALTH,11040
+IDENTIFIER,2941
+LOCATION,17263
+MISC,8374
+PERSON,23790
+SECRET,832
+TEMPORAL,8926

entities_by_type.csv

@@ -0,0 +1,57 @@
+entity_type,entity_group,entities
+NAME_GIVEN,PERSON,17667
+NAME_FAMILY,PERSON,6117
+NAME_ALIAS,PERSON,6
+DATE_OF_BIRTH,TEMPORAL,518
+DATE,TEMPORAL,8340
+AGE,TEMPORAL,37
+PASSPORT_EXPIRATION,TEMPORAL,6
+CREDIT_CARD_EXPIRATION,TEMPORAL,25
+EMAIL_ADDRESS,CONTACT,1280
+PHONE_NUMBER,CONTACT,3376
+LOCATION_ADDRESS,LOCATION,3003
+LOCATION_CITY,LOCATION,7620
+LOCATION_ZIP,LOCATION,1266
+LOCATION_OTHER,LOCATION,5327
+LOCATION_COORDINATES,LOCATION,47
+BANK_ACCOUNT,FINANCIAL,340
+CREDIT_CARD,FINANCIAL,104
+SSN,IDENTIFIER,391
+PASSPORT_NUMBER,IDENTIFIER,20
+INTERNAL_ID,IDENTIFIER,1811
+IMEI,IDENTIFIER,125
+IMSI,IDENTIFIER,120
+RIO,IDENTIFIER,95
+HEALTH_INSURANCE_ID,IDENTIFIER,32
+DRIVER_LICENSE_NUMBER,IDENTIFIER,8
+STUDENT_ID,IDENTIFIER,12
+INSURANCE_ID,IDENTIFIER,24
+REGISTRATION_NUMBER,IDENTIFIER,232
+VEHICLE_REGISTRATION_NUMBER,IDENTIFIER,17
+SERIAL_NUMBER,IDENTIFIER,2
+RESERVATION_NUMBER,IDENTIFIER,8
+TRANSACTION_ID,IDENTIFIER,10
+VISA_NUMBER,IDENTIFIER,7
+BIRTH_CERTIFICATE_NUMBER,IDENTIFIER,5
+TRAIN_TICKET_NUMBER,IDENTIFIER,6
+IDENTITY_DOCUMENT_NUMBER,IDENTIFIER,16
+USERNAME,DIGITAL,170
+URL,DIGITAL,173
+IP_OR_MAC,DIGITAL,167
+ORGANIZATION,CONTEXTUAL,8840
+PRODUCT,CONTEXTUAL,2196
+TAX_ID,SECRET,252
+PIN,SECRET,466
+PUK,SECRET,3
+CVV,SECRET,8
+PASSWORD,SECRET,85
+SIM_CODE,SECRET,1
+ACTIVATION_CODE,SECRET,9
+VERIFICATION_CODE,SECRET,7
+AUTH_TOKEN,SECRET,1
+CURRENCY_AMOUNT,MISC,1230
+MEASUREMENT,MISC,5909
+RATIO,MISC,1235
+MEDICAL_CONDITION,HEALTH,4089
+MEDICATION,HEALTH,6943
+BLOOD_TYPE,HEALTH,8

gliner/gliner_train.jsonl

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:bd8c243d9fbfaf5adebd1d966fad10f62bfcf94abee62cb242cfd3c63e086782
+size 11545854

gliner/gliner_train.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ec47407f7c090aaca8f772ee09100d44ebed86db00a0dedb7cff2c3b9816b337
+size 2712940

gliner25/gliner25_train.jsonl

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e0b35a2c441bd83c548ba3a62332cca3723e45e1eb44146c997c0c0ed4d5e499
+size 18704192

gliner25/gliner25_train.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:33e46aad6acbbe83e315e3534ce58876f0469000e9c5861323bf7907854aed6b
+size 3642073

label_registry.json

@@ -0,0 +1,226 @@
+{
+  "NAME_GIVEN": {
+    "entity_group": "PERSON",
+    "description": "Given or personal name of a person. Titles, honorifics and professional roles are excluded."
+  },
+  "NAME_FAMILY": {
+    "entity_group": "PERSON",
+    "description": "Surname or family name of a person. Titles and honorifics are excluded."
+  },
+  "NAME_ALIAS": {
+    "entity_group": "PERSON",
+    "description": "Explicit nickname, alias, pseudonym, or alternate personal name."
+  },
+  "DATE_OF_BIRTH": {
+    "entity_group": "TEMPORAL",
+    "description": "Calendar date or year explicitly referring to a person's birth."
+  },
+  "DATE": {
+    "entity_group": "TEMPORAL",
+    "description": "General calendar date. Clock times are excluded."
+  },
+  "AGE": {
+    "entity_group": "TEMPORAL",
+    "description": "Explicit age of a person."
+  },
+  "PASSPORT_EXPIRATION": {
+    "entity_group": "TEMPORAL",
+    "description": "Date explicitly representing passport expiration or end of validity."
+  },
+  "CREDIT_CARD_EXPIRATION": {
+    "entity_group": "TEMPORAL",
+    "description": "Expiration date of a credit card."
+  },
+  "EMAIL_ADDRESS": {
+    "entity_group": "CONTACT",
+    "description": "Email address."
+  },
+  "PHONE_NUMBER": {
+    "entity_group": "CONTACT",
+    "description": "Telephone, mobile, or fax number."
+  },
+  "LOCATION_ADDRESS": {
+    "entity_group": "LOCATION",
+    "description": "Physical street, building, house, unit, apartment, or equivalent address."
+  },
+  "LOCATION_CITY": {
+    "entity_group": "LOCATION",
+    "description": "City, town, village, municipality, or locality, including inflected Slovene geographic forms."
+  },
+  "LOCATION_ZIP": {
+    "entity_group": "LOCATION",
+    "description": "Postal or ZIP code."
+  },
+  "LOCATION_OTHER": {
+    "entity_group": "LOCATION",
+    "description": "Country, region, state, continent, landmark, or another geographic entity not covered by address, city, or postal code."
+  },
+  "LOCATION_COORDINATES": {
+    "entity_group": "LOCATION",
+    "description": "Latitude, longitude, GPS, or other explicit geographic coordinates."
+  },
+  "BANK_ACCOUNT": {
+    "entity_group": "FINANCIAL",
+    "description": "IBAN or another bank-account identifier."
+  },
+  "CREDIT_CARD": {
+    "entity_group": "FINANCIAL",
+    "description": "Full, masked, or explicitly identified partial credit-card number."
+  },
+  "SSN": {
+    "entity_group": "IDENTIFIER",
+    "description": "Social-security or unique national personal identifier, including an explicitly identified Slovenian EMŠO."
+  },
+  "PASSPORT_NUMBER": {
+    "entity_group": "IDENTIFIER",
+    "description": "Passport number or passport identifier."
+  },
+  "INTERNAL_ID": {
+    "entity_group": "IDENTIFIER",
+    "description": "Internal client, account, order, invoice, record, case, file, or system identifier where no more specific identifier type applies."
+  },
+  "IMEI": {
+    "entity_group": "IDENTIFIER",
+    "description": "International Mobile Equipment Identity device identifier."
+  },
+  "IMSI": {
+    "entity_group": "IDENTIFIER",
+    "description": "International Mobile Subscriber Identity identifier."
+  },
+  "RIO": {
+    "entity_group": "IDENTIFIER",
+    "description": "Telecommunications number-portability RIO identifier."
+  },
+  "HEALTH_INSURANCE_ID": {
+    "entity_group": "IDENTIFIER",
+    "description": "Health-insurance, health-card, or healthcare-beneficiary identifier."
+  },
+  "DRIVER_LICENSE_NUMBER": {
+    "entity_group": "IDENTIFIER",
+    "description": "Driver's licence identifier or licence number."
+  },
+  "STUDENT_ID": {
+    "entity_group": "IDENTIFIER",
+    "description": "Student identification number."
+  },
+  "INSURANCE_ID": {
+    "entity_group": "IDENTIFIER",
+    "description": "Insurance identifier or policy identifier other than a health-insurance identifier."
+  },
+  "REGISTRATION_NUMBER": {
+    "entity_group": "IDENTIFIER",
+    "description": "Explicit registration, register, licence, or evidential identifier where no more specific identifier type applies."
+  },
+  "VEHICLE_REGISTRATION_NUMBER": {
+    "entity_group": "IDENTIFIER",
+    "description": "Vehicle registration identifier or licence-plate number."
+  },
+  "SERIAL_NUMBER": {
+    "entity_group": "IDENTIFIER",
+    "description": "Explicit serial number of a device, product, or document."
+  },
+  "RESERVATION_NUMBER": {
+    "entity_group": "IDENTIFIER",
+    "description": "Booking or reservation identifier."
+  },
+  "TRANSACTION_ID": {
+    "entity_group": "IDENTIFIER",
+    "description": "Transaction or payment identifier."
+  },
+  "VISA_NUMBER": {
+    "entity_group": "IDENTIFIER",
+    "description": "Visa identifier or visa number."
+  },
+  "BIRTH_CERTIFICATE_NUMBER": {
+    "entity_group": "IDENTIFIER",
+    "description": "Birth-certificate identifier or certificate number."
+  },
+  "TRAIN_TICKET_NUMBER": {
+    "entity_group": "IDENTIFIER",
+    "description": "Train-ticket identifier or ticket number."
+  },
+  "IDENTITY_DOCUMENT_NUMBER": {
+    "entity_group": "IDENTIFIER",
+    "description": "Number of an identity document when no more specific document identifier type applies."
+  },
+  "USERNAME": {
+    "entity_group": "DIGITAL",
+    "description": "Account name, login name, username, or user handle."
+  },
+  "URL": {
+    "entity_group": "DIGITAL",
+    "description": "Web URL or web link."
+  },
+  "IP_OR_MAC": {
+    "entity_group": "DIGITAL",
+    "description": "IPv4 address, IPv6 address, or MAC address."
+  },
+  "ORGANIZATION": {
+    "entity_group": "CONTEXTUAL",
+    "description": "Named company, institution, court, ministry, municipality acting as an organization, authority, agency, university, association, hospital, or other organization."
+  },
+  "PRODUCT": {
+    "entity_group": "CONTEXTUAL",
+    "description": "Named proprietary or branded product, service, application, software, or device. Generic product categories are excluded."
+  },
+  "TAX_ID": {
+    "entity_group": "SECRET",
+    "description": "Tax number or tax identifier."
+  },
+  "PIN": {
+    "entity_group": "SECRET",
+    "description": "Personal identification number or explicitly identified PIN code."
+  },
+  "PUK": {
+    "entity_group": "SECRET",
+    "description": "Personal unblocking key or explicitly identified PUK code."
+  },
+  "CVV": {
+    "entity_group": "SECRET",
+    "description": "Credit-card verification or security code."
+  },
+  "PASSWORD": {
+    "entity_group": "SECRET",
+    "description": "Password or passphrase."
+  },
+  "SIM_CODE": {
+    "entity_group": "SECRET",
+    "description": "Explicit SIM-related security code not better represented as PIN or PUK."
+  },
+  "ACTIVATION_CODE": {
+    "entity_group": "SECRET",
+    "description": "Code explicitly used to activate an account, device, SIM, product, or service."
+  },
+  "VERIFICATION_CODE": {
+    "entity_group": "SECRET",
+    "description": "One-time verification or confirmation code, including SMS verification codes."
+  },
+  "AUTH_TOKEN": {
+    "entity_group": "SECRET",
+    "description": "Authentication or access token such as a JWT, bearer token, or access-token value."
+  },
+  "CURRENCY_AMOUNT": {
+    "entity_group": "MISC",
+    "description": "Explicit monetary amount represented by a numeric value with currency or clearly established monetary context."
+  },
+  "MEASUREMENT": {
+    "entity_group": "MISC",
+    "description": "Numeric quantity with a physical, scientific, medical, technical, storage, speed, dosage, concentration, or duration unit."
+  },
+  "RATIO": {
+    "entity_group": "MISC",
+    "description": "Percentage, proportion, or ratio."
+  },
+  "MEDICAL_CONDITION": {
+    "entity_group": "HEALTH",
+    "description": "Explicit disease, diagnosis, disorder, syndrome, injury, or clinical condition."
+  },
+  "MEDICATION": {
+    "entity_group": "HEALTH",
+    "description": "Medicine, drug, pharmaceutical preparation, active medicinal substance, or therapeutic drug name."
+  },
+  "BLOOD_TYPE": {
+    "entity_group": "HEALTH",
+    "description": "Explicit blood group or blood type."
+  }
+}

rows_by_domain.csv

@@ -0,0 +1,9 @@
+domain,rows
+banking,94
+finance,82
+general,8742
+healthcare,78
+legal,1065
+medical,1000
+news,2125
+telecom,1994

rows_by_source.csv

@@ -0,0 +1,9 @@
+source,rows
+COLESLAW,1000
+PoVeJMo-VeMo-Med,1000
+e3-jsi,423
+eeisenzopf,1494
+gigafida,2125
+kurkowski,1026
+tanaos,7612
+telekom_synthetic,500