Datasets:

Hyukkyu
/

train-law-stackexchange

pretty_name: Training · Law Stack Exchange
license: cc-by-sa-4.0
language:
  - en
multilinguality:
  - monolingual
task_categories:
  - text-retrieval
task_ids:
  - document-retrieval
tags:
  - train
  - retrieval
  - legal
configs:
  - config_name: corpus
    data_files:
      - split: train
        path: corpus/train-*.parquet
  - config_name: hard-negatives
    data_files:
      - split: train
        path: hard-negatives/train-*.parquet
  - config_name: qrels
    data_files:
      - split: train
        path: qrels/train-*.parquet
  - config_name: queries
    data_files:
      - split: train
        path: queries/train-*.parquet
  - config_name: teacher-scores
    data_files:
      - split: train
        path: teacher-scores/train-*.parquet

Law Stack Exchange — Training, unified schema

A seeded sample of ymoslem/Law-StackExchange, made into retrieval training pairs and reshaped into the strict schema shared by every dataset in this collection. One of the 15 domain sources (code, medical, science, finance, legal) added to the collection's general sources.


Source	`ymoslem/Law-StackExchange` @ `ab2dbaad9a71`
Task	legal question → answer
Domain · languages	legal · eng
Queries / documents / qrels	24,326 / 24,330 / 24,326
Qrels per query	min 1 · mean 1.0 · max 1
Score values	2 ×24,326 (2: the first positive, 1: any other)
Layout	`queries` · `corpus` · `qrels` · `hard-negatives` · `teacher-scores`, split `train`
Splits	`corpus`: train · `hard-negatives`: train · `qrels`: train · `queries`: train · `teacher-scores`: train
Hard negatives	sources: `dense` · 2,409,643 rows
Teacher scores	none yet (0 rows): `jinaai/jina-reranker-v3.5` scores come next
Ids	`sha1(text)[:20]`; identical texts collapse to one document / query
License	`cc-by-sa-4.0`

Schema

config	columns	rules
`queries`	`id: string`, `text: string`	ids unique and non-empty; every query has ≥ 1 qrel
`corpus`	`id: string`, `title: string`, `text: string`	`title` is always present (`""` when the source has none)
`qrels`	`query-id: string`, `corpus-id: string`, `score: int32`	referential integrity to both tables; no duplicate pairs; no floats
`hard-negatives`	`query-id: string`, `corpus-id: string`, `rank: int32`, `source: string`	one row per negative; `(query-id, corpus-id, source)` unique; never a labelled positive of the same query
`teacher-scores`	`query-id: string`, `corpus-id: string`, `teacher: string`, `score: float32`	one row per scored pair (positives included); a row means scored — never a placeholder

Files are Parquet, sorted by id, zstd-compressed, sharded at 500 MB. Every rule above is checked before publishing; provenance.json records the source revision, what changed, and the output file hashes.

What changed from the source

sampled: a seeded random sample (seed 1) of up to 25,000 pairs
reshaped: the question title and body (HTML removed, at most 2,000 characters) are the query, the highest-scored answer (HTML removed, at least 40 characters) the document
decontaminated (exact): a pair was dropped when its normalised query equals any evaluation query, or a positive equals a document of a test or dev corpus; a repeated query keeps its first pair
decontaminated (near-duplicates): 5 passages that nearly copy an evaluation document some evaluation query judges relevant, and 4 queries that nearly copy an evaluation query (word 13-grams for passages, 8-grams for queries; at least half shared with one text of the 23 test sets (BEIR, RTEB, LitSearch) or the 6 dev sets) were removed, and with them 9 queries in total; near copies of evaluation-corpus documents that no evaluation query judges relevant were kept
text: leading and trailing whitespace stripped; otherwise as converted above
ids re-keyed to sha1(text)[:20]: 0 documents and 0 queries collapsed into identical texts
added a title column filled with "" (the source has none)

Hard negatives and teacher scores

Filled by the collection's annotation pipeline (annotation=jina35). Interim: the candidates are final, the teacher scores are still to come.

Candidates: dense retrieval with jinaai/jina-embeddings-v5-text-small over this corpus to depth 1,000; 100 candidates per query drawn from the rank windows 1–30 (30), 31–100 (30), 101–300 (20), 301–1000 (20), the query's labelled positives excluded. rank is the dense rank; source is dense for a mined row and dataset for a negative the source labels itself.
Teacher scores: none yet. teacher-scores holds 0 rows until the jinaai/jina-reranker-v3.5 scores (listwise, as in the other repositories) are filled in; datasets cannot return a 0-example split, so read that file with pyarrow / pandas meanwhile. The candidates stay.

configs	queries	hard negatives	teacher scores
`hard-negatives` · `teacher-scores`	24,326 (all)	2,409,643 (2,409,643 dense)	0

Load it

from datasets import load_dataset
queries   = load_dataset("Hyukkyu/train-law-stackexchange", "queries", split="train")
corpus    = load_dataset("Hyukkyu/train-law-stackexchange", "corpus", split="train")
qrels     = load_dataset("Hyukkyu/train-law-stackexchange", "qrels", split="train")
negatives = load_dataset("Hyukkyu/train-law-stackexchange", "hard-negatives", split="train")
scores    = load_dataset("Hyukkyu/train-law-stackexchange", "teacher-scores", split="train")

License and attribution

The data is redistributed under the source's terms — cc-by-sa-4.0. All credit belongs to the original authors; see the source repository (https://huggingface.co/datasets/ymoslem/Law-StackExchange). This repository is an independent repackaging.

Share-alike. The source is CC BY-SA 4.0, and so is this copy.