Document 80-sample public CN/EU release

README.md

@@ -27,53 +27,44 @@ configs:
     path: data/test.parquet
 dataset_info:
   features:
-  - name: id
-    dtype: int64
-  - name: query_img_id
-    dtype: string
   - name: query_pn
     dtype: string
   - name: query_img_path
     dtype: image
   - name: target_pns
     sequence: string
-  - name: target_img_ids
-    sequence: string
-  - name: pair_name
-    dtype: string
   - name: picture_type
     dtype: string
-  - name: one_level_loc
+  - name: loc
     dtype: string
-  - name: two_level_loc
+  - name: sub_loc
     dtype: string
-  - name: country
+  - name: jurisdiction
     sequence: string
-  - name: version
-    dtype: string
+  - name: is_public
+    dtype: int64
   splits:
   - name: test
-    num_examples: 91
+    num_examples: 80
 ---
 
 # PatSnap Design FTO Bench
 
-A Bench for evaluating **design patent Freedom-To-Operate (FTO)** retrieval systems on cross-modal image search. Each sample provides a query product image (or design patent figure) plus the ground truth set of target design patents that constitute infringement risk, as confirmed by patent invalidation proceedings.
+This dataset evaluates **design patent Freedom-To-Operate (FTO) image retrieval**. Given a query image, a system ranks design patents; `target_pns` contains the annotated target publication numbers. The public test split contains 80 invalidation-proceeding records selected with `is_public = 1`. It contains no e-commerce complaint samples.
 
-> 🐙 **GitHub mirror:** This dataset is also published as part of the [`patsnap/patent-bench`](https://github.com/patsnap/patent-bench/tree/main/design-fto-bench) monorepo, where you can find the reference metric scripts (`search_metrics.py`) and additional sub-Benches.
+The 80 query PNGs are embedded in `data/test.parquet`. The `data/image/` files in the repository belong to an earlier release and are not the source for the current test split.
 
 ## Dataset Overview
 
 | Property | Value |
 |----------|-------|
-| **Total samples** | 91 |
-| **Source** | Real patent invalidation proceedings |
-| **Jurisdictions** | CN (100% for the released subset) |
-| **Modality** | PNG images (product photo ↔ patent line drawing / photo) |
-| **Image directory** | `data/image/<jurisdiction>/<class>/<sub>/<pn>/<file>.png` (91 PNG files, ~6.8 MB) |
-| **Ground truth** | Patent pairs confirmed as infringement-equivalent through patent invalidation proceedings |
-| **Locarno (LOC) coverage** | All 26 first-level LOC classes |
-| **License** | CC BY-NC 4.0 |
+| Public test samples | 80 |
+| Jurisdictions (`jurisdiction`) | CN: 40; EU: 40 |
+| Query images | 80 embedded PNGs |
+| Ground truth | 98 target-PN entries across 80 samples |
+| Locarno coverage | 20 distinct first-level codes after splitting `OR` combinations |
+| Source | Invalidation proceedings |
+| License | CC BY-NC 4.0 |
 
 ## Quick Start
 
@@ -81,79 +72,39 @@ A Bench for evaluating **design patent Freedom-To-Operate (FTO)** retrieval syst
 from datasets import load_dataset
 
 ds = load_dataset("PatSnap/design-fto-bench", split="test")
-print(f"Total samples: {len(ds)}")
+print(len(ds))  # 80
 
-# Inspect one sample
 sample = ds[0]
-print(sample["query_pn"], sample["pair_name"])
-
-# query_img_path is a PIL Image (bytes embedded in the Parquet, no external lookup needed)
-img = sample["query_img_path"]
-print(f"Query image: {img.size}, mode={img.mode}")
-
-# Targets are the set of design patents whose images constitute infringement risk
-print(sample["target_pns"], sample["target_img_ids"])
+image = sample["query_img_path"]  # PIL image; bytes are embedded in Parquet
+print(image.size, sample["target_pns"], sample["jurisdiction"])
 ```
 
+`query_pn` is `null` for 27 samples. Use the query image as the input; do not infer a missing query PN from its filename or target PNs.
+
 ## Data Fields
 
 | Field | Type | Description |
 |-------|------|-------------|
-| `id` | int64 | Sample identifier |
-| `query_img_id` | string | Identifier of the query image |
-| `query_pn` | string | Publication number of the query patent (PatSnap standardized PN) |
-| `query_img_path` | string | Relative path to the query image under `data/image/` |
-| `target_pns` | list[string] | Ground truth target design-patent PNs that constitute infringement risk |
-| `target_img_ids` | list[string] | Image identifiers of the target patents |
-| `pair_name` | string | Pair identifier from the invalidation proceeding |
-| `picture_type` | string | Source of the GT pair (e.g. 无效 = invalidation proceeding) |
-| `one_level_loc` | string | First-level Locarno classification code |
-| `two_level_loc` | string | Second-level Locarno classification code (e.g. 14-03) |
-| `country` | list[string] | Country/jurisdiction codes of the sample |
-| `version` | string | Dataset version (e.g. 1.1) |
-
-## How to Use the Query
+| `query_pn` | nullable string | Query patent publication number, if available |
+| `query_img_path` | image | Query PNG embedded in the Parquet file; its stored path is relative to the source package's `data/` directory |
+| `target_pns` | list[string] | Annotated target patent publication numbers, in source order |
+| `picture_type` | string | `Invalidation Proceedings` for every published sample |
+| `loc` | string | First-level Locarno code; may contain `OR` |
+| `sub_loc` | string | More specific Locarno code; may contain `OR` |
+| `jurisdiction` | list[string] | Target search jurisdiction (`["CN"]` or `["EU"]`) |
+| `is_public` | int64 | `1` for every published sample |
 
-The query input is the **query product image** at `data/image/<query_img_path>`. Each sample's `target_pns` lists the design patents that an FTO retrieval system should return.
+## Evaluation
 
-## Evaluation Metrics
+Use `target_pns` as the patent-number ground truth. Hit Rate @ K is the fraction of all 80 test samples with at least one target PN in the top K results. PRES @ N also accounts for the ranks of retrieved targets; the usual depth is N = 200. Missing or failed queries remain in the 80-sample denominator for a full-set evaluation. Report the dataset revision, ranking cutoff, PN normalization rules, and treatment of query patents that also appear in `target_pns`.
 
-| Metric | Description |
-|--------|-------------|
-| **Hit Rate @ K** | % of samples with ≥1 GT patent in top K (K = 10, 50, 100, 200) |
-| **PRES @ N** | Patent Retrieval Evaluation Score (Magdy & Jones 2010, with miss-penalty correction): single score in `[0, 1]` jointly capturing how many GT patents are retrieved within top-N and how highly they are ranked. PRES = 1.0 means every GT patent appears at the top; PRES = 0 means none are found within N. Default N = 200. |
+The [reference metric functions](https://github.com/patsnap/patent-bench/blob/main/common/metrics/search_metrics.py) define Hit Rate and PRES. Their command-line workflow expects a JSONL file with an `id` field; this public Parquet intentionally has no internal sample IDs. Use the functions with ranked results aligned to the 80 published rows, or adapt the CLI input locally. Image-level target IDs are not part of this public schema.
 
-The reference metric scripts (with `strict / leaderboard` mode by default and ranked-list schema validation) are available in the [`patsnap/patent-bench`](https://github.com/patsnap/patent-bench/blob/main/common/metrics/search_metrics.py) GitHub repo.
+## Release Notes and Limitations
 
-### Scoring Grades (Hit Rate @ Top@100)
-
-| Grade | Hit Rate | Description |
-|-------|----------|-------------|
-| **A** | ≥ 90% | Excellent — suitable for direct professional use |
-| **B** | ≥ 75% | Good — effective as a high-efficiency screening tool |
-| **C** | ≥ 60% | Acceptable — requires human review of key results |
-| **D** | < 60% | Below standard — model improvement needed |
-
-## Distribution
-
-### By Jurisdiction
-
-| Jurisdiction | Count | Percentage |
-|--------------|-------|------------|
-| CN | 91 | 100% |
-
-> The v1.1 public release contains only invalidation-proceeding samples (CN). Future releases (v2) will incorporate cross-jurisdiction TRO data (US/EP/JP).
-
-### By Locarno Classification
-
-Coverage spans all 26 first-level LOC classes.
-
-## Limitations
-
-- **Retrieval-only Bench**: Evaluates the search/retrieval step only; does not cover infringement adjudication or court-ruling outcomes.
-- **GT based on invalidation proceedings**: This subset (v1.1) is restricted to CN invalidation-proceeding pairs. E-commerce infringement-complaint samples are retained internally for client confidentiality.
-- **Single-jurisdiction**: CN only in this release.
-- **Visual similarity ≠ legal infringement**: A retrieval system returning a top-1 hit does not constitute a legal infringement determination; results are inputs to professional FTO review.
+- The 2026-10-13 public selection replaces the earlier 91-row CN-only test split with 80 rows across CN and EU. Its schema also changed. Results on the earlier release are not directly comparable; pin a dataset commit when reporting scores.
+- This benchmark covers retrieval, not infringement adjudication or legal FTO conclusions. Visual similarity alone is not a legal finding.
+- The public selection contains only invalidation-proceeding records. It is not a representative sample of all product designs, jurisdictions, or FTO matters.
 
 ## Citation
 
@@ -163,16 +114,14 @@ Coverage spans all 26 first-level LOC classes.
   author = {PatSnap},
   year   = {2026},
   url    = {https://huggingface.co/datasets/PatSnap/design-fto-bench},
-  note   = {A Bench for evaluating design-patent freedom-to-operate image-retrieval systems}
+  note   = {A benchmark for design-patent FTO image retrieval}
 }
 ```
 
 ## License
 
-Released under [CC BY-NC 4.0](https://creativecommons.org/licenses/by-nc/4.0/) — research and non-commercial evaluation purposes only.
+Released under [CC BY-NC 4.0](https://creativecommons.org/licenses/by-nc/4.0/) for research and non-commercial evaluation.
 
 ## Try the Production System
 
-Experience the **PatSnap Design FTO AI Agent** — the commercial system referenced in this Bench.
-
-🔗 [Try it on PatSnap Eureka](https://eureka.patsnap.com/ip/checking/?from=benchmark_huggingface#/design-fto)
+[PatSnap Design FTO on Eureka](https://eureka.patsnap.com/ip/checking/?from=benchmark_huggingface#/design-fto)