five

albertgd/legal-divorce-es

收藏
Hugging Face2026-03-26 更新2026-03-29 收录
下载链接:
https://hf-mirror.com/datasets/albertgd/legal-divorce-es
下载链接
链接失效反馈
官方服务:
资源简介:
--- language: - es license: cc-by-4.0 task_categories: - question-answering - text-generation - summarization tags: - legal - spanish - divorce - constitutional-court - rag - tribunal-constitucional - amparo - family-law pretty_name: Spanish Constitutional Court — Divorce & Family Law Cases size_categories: - n<1K --- # Spanish Constitutional Court — Divorce & Family Law Cases A curated dataset of **168 rulings** from the Spanish Constitutional Court (*Tribunal Constitucional de España*) related to divorce, separation, and family law proceedings. Each record contains the full anonymised case text, structured legal sections, and agent-ready distilled learnings — designed for building RAG pipelines and fine-tuning legal AI assistants in Spanish. ## Dataset Details ### Source Cases retrieved from the [official HJ search engine](https://hj.tribunalconstitucional.es/) of the *Tribunal Constitucional de España*. All source documents are public legal records. ### Coverage | Attribute | Value | |---|---| | Total cases | 168 | | Decision type — SENTENCIA | 97 | | Decision type — AUTO | 71 | | Date range | 1986 – present | | Language | Spanish (es) | | Size | ~3.8 MB (JSONL) | ### PII Handling All personal names of private individuals have been anonymised using [Microsoft Presidio](https://github.com/microsoft/presidio) with the `es_core_news_lg` spaCy model: - **Person names** → `[PERSONA_1]`, `[PERSONA_2]`, … (numbered consistently within each document so the narrative remains coherent) - **DNI / NIE** → `[DNI]` / `[NIE]` - **Phone numbers** → `[TELEFONO]` - **Email addresses** → `[EMAIL]` Court names, institutional parties, and judges are **preserved** as they are essential legal context and already part of the public record. --- ## Dataset Structure ### Fields | Field | Type | Description | |---|---|---| | `case_id` | string | Internal identifier, e.g. `CASO_11051` | | `title` | string | Full official title of the ruling | | `ecli` | string | European Case Law Identifier, e.g. `ECLI:ES:TC:1986:1031A` | | `url` | string | Source URL on the Tribunal Constitucional website | | `referencia` | string | Short reference, e.g. `AUTO 1031/1986` | | `tipo` | string | `SENTENCIA` (judgment) or `AUTO` (court order) | | `fecha` | string | Date of the ruling in Spanish long form | | `hechos` | string | Facts / antecedents section (anonymised) | | `doctrina` | string | Legal doctrine and reasoning section (anonymised) | | `fallo` | string | Operative part / ruling (anonymised) | | `sintesis` | string | Official descriptive and analytical synthesis | | `learnings` | string | Agent-ready distilled learnings: key doctrine, takeaways, and practical notes | | `text` | string | Full composite text combining all sections — ready for embedding or fine-tuning | ### Example Record ```python { "case_id": "CASO_1201", "title": "Sala Segunda. SENTENCIA 260/1988, de 22 de diciembre", "ecli": "ECLI:ES:TC:1988:260", "tipo": "SENTENCIA", "fecha": "22 de diciembre de 1988", "hechos": "El Procurador de los Tribunales don [PERSONA_3], en nombre de doña [PERSONA_2], interpone recurso de amparo...", "learnings": "## Doctrina que se fija\n- La Disposición adicional décima de la Ley 30/1981...", "text": "# Sala Segunda. SENTENCIA 260/1988...\n\n## Hechos\n..." } ``` --- ## Usage ### Load the dataset ```python from datasets import load_dataset ds = load_dataset("albertgd/legal-divorce-es") print(ds["train"]) ``` ### RAG pipeline (embed + retrieve) ```python from datasets import load_dataset from sentence_transformers import SentenceTransformer import numpy as np ds = load_dataset("albertgd/legal-divorce-es", split="train") model = SentenceTransformer("sentence-transformers/paraphrase-multilingual-mpnet-base-v2") embeddings = model.encode(ds["text"], show_progress_bar=True) # Query query = "¿Cuándo puede modificarse una pensión compensatoria tras el divorcio?" q_emb = model.encode([query]) scores = np.dot(embeddings, q_emb.T).squeeze() top_idx = scores.argsort()[-3:][::-1] for i in top_idx: print(ds[int(i)]["title"]) print(ds[int(i)]["learnings"][:300]) print() ``` ### Fine-tuning (instruction format) ```python from datasets import load_dataset ds = load_dataset("albertgd/legal-divorce-es", split="train") def to_instruction(example): return { "prompt": f"Analiza el siguiente caso legal español y extrae los aprendizajes clave:\n\n{example['text']}", "response": example["learnings"] } ft_ds = ds.map(to_instruction) ``` --- ## Legal Context The *Tribunal Constitucional de España* is Spain's supreme interpreter of the Constitution. It handles: - **Recurso de amparo** — constitutional protection appeals by individuals against violations of fundamental rights - **Cuestiones de inconstitucionalidad** — questions of unconstitutionality raised by courts In the context of divorce law, these cases primarily concern: - Application of **Ley 30/1981** (divorce reform) — retroactivity, transitional provisions - **Art. 14 CE** — equality / non-discrimination in family settlements - **Art. 24 CE** — right to effective judicial protection - Enforcement of foreign divorce judgments (*exequátur*) - Pension and property disputes in separation / divorce proceedings --- ## Citation If you use this dataset, please cite the source: ```bibtex @dataset{albertgd_legal_divorce_es_2026, author = {albertgd}, title = {Spanish Constitutional Court — Divorce \& Family Law Cases}, year = {2026}, publisher = {Hugging Face}, url = {https://huggingface.co/datasets/albertgd/legal-divorce-es}, note = {Cases sourced from the Tribunal Constitucional de España (hj.tribunalconstitucional.es). PII anonymised with Microsoft Presidio.} } ``` Original case texts: © Tribunal Constitucional de España — public records, freely accessible at [hj.tribunalconstitucional.es](https://hj.tribunalconstitucional.es/). --- ## License [CC BY 4.0](https://creativecommons.org/licenses/by/4.0/) — you may share and adapt the dataset with attribution.
提供机构:
albertgd
5,000+
优质数据集
54 个
任务类型
进入经典数据集
二维码
社区交流群

面向社区/商业的数据集话题

二维码
科研交流群

面向高校/科研机构的开源数据集话题

数据驱动未来

携手共赢发展

商业合作