AmirMohseni/WildChat-Legal-Classification-V2-LegalOnly

WildChat Legal Classification (Legal Only)

English multi-turn conversations from WildChat-1M where the user seeks legal guidance (seeks_legal_guidance=true), silver-labeled with GPT-5.4. Each row includes the full conversation, classification fields, and a nested list of incidents — distinct legal-guidance questions extracted from the thread.

Use this dataset when you need legal-guidance rows only (no non-legal negatives), or when you want conversation-level access to extracted questions before flattening.

For one row per question, see AmirMohseni/WildChat-Legal-Questions.

Related datasets

Dataset	Contents
AmirMohseni/WildChat-Legal-Classification-V2-Balanced	Equal legal-guidance / non-guidance rows for detection + topic tasks
AmirMohseni/WildChat-Legal-Questions	Flattened one-row-per-question view derived from incidents

Splits

Each split is available as a config key (total, train, val, test). Partition ratio: 80 / 10 / 10 (shuffle seed 42).

Split	Examples
total	1,061
train	848
val	106
test	107

Schema

Conversation-level columns

Same fields as the balanced dataset:

Column	Type	Notes
id	string	WildChat conversation id
conversation	list of {role, content}	Multi-turn user–assistant thread
seeks_legal_guidance	bool	Always true in this dataset
primary_topic	string	Single best topic for the conversation-level label
topics	list[string]	1–3 reasonable topic labels
contains_legal_content	bool
personalized_or_operational	bool
uncertainty	string	low, medium, or high
relevant_user_turn_ids	list[int]
relevant_assistant_turn_ids	list[int]

incidents column

List of extracted legal-guidance questions within the conversation (may be empty if extraction found none):

Field	Type	Description
user_turn_ids	list[int]	User turn indices for this incident
assistant_turn_ids	list[int]	Assistant turns that directly answer the question
standalone_question_original	string	Question rewritten in the original language
standalone_question_en	string	English rewrite of the standalone question
topics	list[string]	1–3 topic labels for this incident
primary_topic	string	Best topic for this incident

Long chats may contain multiple incidents; each incident can carry its own topic labels.

Usage

from datasets import load_dataset

ds = load_dataset("AmirMohseni/WildChat-Legal-Classification-V2-LegalOnly", split="train")
row = ds[0]
for inc in row["incidents"]:
    print(inc["standalone_question_en"], "->", inc["primary_topic"])

Dataset construction

Source

Conversations are drawn from allenai/WildChat-1M. WildChat is released under ODC-BY; this derivative adds GPT-5.4 silver labels under CC-BY-4.0.

Labeling pipeline

All annotations use GPT-5.4 with structured JSON outputs.

Pass 1 (low effort). GPT-5.4 triages each conversation into legal, non-legal, or maybe legal to narrow the candidate pool (~11k conversations).

Pass 2 (high effort). GPT-5.4 re-labels legal and maybe legal rows with conversation-level fields (contains_legal_content, seeks_legal_guidance, personalized_or_operational, topics, primary_topic, uncertainty, relevant turn ids). Only rows with seeks_legal_guidance=true are retained in this dataset (1,061 conversations).

English filtering. Rows are kept only when user-authored text is English-dominant. GPT-5.4 additionally filters out conversations that appeared English on metadata but were predominantly written in another language.

Incident extraction (high effort). On each legal-guidance conversation, GPT-5.4 identifies distinct user legal questions, merges adjacent turns about the same issue, rewrites each as a standalone question (with English translation when needed), and assigns per-incident topic labels. Results are stored in the incidents list.

A 100-item human audit of GPT-5.4 labels on a manually curated 50/50 sample found ~93% overall agreement with human judgment.

Topic taxonomy

Fourteen labels, kept compact enough for training but separable in practice:

Label	Short description
FAMILY	Marriage, divorce, custody, support, adoption, guardianship, domestic violence.
HOUSING	Rent, eviction, deposits, mortgages, foreclosure, habitability, neighbors, housing subsidies.
WORK	Employment contracts, wages, dismissal, discrimination at work, leave, workplace safety.
PUBLIC_BENEFITS	Unemployment, disability, pensions, welfare, eligibility, appeals on benefits.
CRIMINAL_JUSTICE	Police, arrest, charges, fines, prosecution, defense, victims' rights, criminal procedure.
CONSUMER_DEBT	Purchases, warranties, refunds, scams, debt collection, loans, bankruptcy, credit.
CONTRACTS	Civil agreements; breach/interpretation not better captured by work/housing/consumer/business.
IMMIGRATION	Visas, residence permits, asylum, citizenship, deportation, family migration.
BUSINESS	Company formation, shareholders, commercial compliance, B2B disputes, self-employment.
DATA_PRIVACY	Personal data, surveillance, GDPR, data deletion, consent, platform data practices.
INTELLECTUAL_PROPERTY	Copyright, trademark, patent, trade secrets, licensing, infringement.
CIVIL_RIGHTS	Discrimination outside work/housing, free speech, due process, equal treatment.
INTERNATIONAL_CROSS_BORDER	Choice of law, jurisdiction, treaty-based questions, cross-border enforcement.
OTHER	Genuinely legal but not covered above.

Selection rules applied during labeling:

• Use CONTRACTS only when the issue is mainly about a civil agreement and is not better captured by WORK, HOUSING, CONSUMER_DEBT, or BUSINESS.
• Use INTERNATIONAL_CROSS_BORDER only when the cross-border or jurisdictional aspect is central to the answer, not merely incidental.

Limitations

• Labels are silver annotations from GPT-5.4, not professional legal review. Do not use for legal advice or high-stakes routing without human oversight.
• English-focused; jurisdiction is preserved in text when stated but not normalized.
• Incident extraction may miss or merge questions; see the flattened questions dataset for a simpler per-question view.

Citation

If you use this dataset, please cite WildChat and acknowledge GPT-5.4 silver labeling:

@misc{wildchat-legal-classification-legalonly,
  title={WildChat Legal Classification (Legal Only)},
  author={Mohseni, Amir},
  year={2025},
  publisher={Hugging Face},
  howpublished={\\url{https://huggingface.co/datasets/AmirMohseni/WildChat-Legal-Classification-V2-LegalOnly}}
}