MMG/spanishBFF2
收藏资源简介:
Spanish-BFF-2是使用GPT-4生成的第二个西班牙语词典,包含了76,963个词条及其定义。这些词条涵盖了名词、形容词、动词和副词等词性。数据集旨在促进自然语言处理(NLP)的理解和发展,并推动负责任的使用。该版本尚未经过后处理以减轻AI模型可能生成的错误、偏见或幻觉。
Spanish-BFF-2 is the second Spanish AI-generated dictionary using GPT-4, containing a total of 76,963 lemmas with their definitions. These lemmas correspond to nominal, adjectival, verbal, and adverbial classes. The dataset aims to contribute to a better understanding and development of NLP and promote responsible use. This version has not been postprocessed to mitigate potential errors, biases, or hallucinations the AI model could have generated.
数据集描述
数据集名称: Spanish Built Factual Freectianary 2 (Spanish-BFF-2)
数据集简介: Spanish-BFF-2 是使用 GPT-4 生成的第二个西班牙语 AI 词典。包含 76,963 个词条及其定义,涵盖名词、形容词、动词和副词类别。
语言: 西班牙语 (es)
数据集大小: 10K<n<100K
多语言性: 单语种
标注创建者: AI 生成
许可证: gpl-3.0
数据集结构
数据实例
plaintext "retransmisión= [{"pos_tag": "Nombre femenino", "definition": "Acción y efecto de retransmitir un evento, generalmente por televisión o radio.", "example": "La retransmisión del partido de fútbol fue vista por millones de personas."}]
数据字段
plaintext { pos_tag: str definition: str example: str }
数据分割
| 分割 | 大小 |
|---|---|
train |
76,963 |
使用数据集的注意事项
数据集的社会影响
该语料库是第二个由 LLM 生成的开源完整词典。旨在促进对 NLP 的更好理解和开发,并推广负责任的使用。
偏见和幻觉
此版本尚未经过后处理以减轻 AI 模型可能生成的潜在错误、偏见或幻觉。
引用
plaintext @misc{ortegamartín2024building, title={Building another Spanish dictionary, this time with GPT-4}, author={Miguel Ortega-Martín and Óscar García-Sierra and Alfonso Ardoiz and Juan Carlos Armenteros and Ignacio Garrido and Jorge Álvarez and Camilo Torrón and Iñigo Galdeano and Ignacio Arranz and Oleg Vorontsov and Adrián Alonso}, year={2024}, eprint={2406.11218}, archivePrefix={arXiv}, primaryClass={id=cs.CL full_name=Computation and Language is_active=True alt_name=cmp-lg in_archive=cs is_general=False description=Covers natural language processing. Roughly includes material in ACM Subject Class I.2.7. Note that work on artificial languages (programming languages, logics, formal systems) that does not explicitly address natural-language issues broadly construed (natural-language processing, computational linguistics, speech, text retrieval, etc.) is not appropriate for this area.} }



