A French corpus annotated for multiword expressions with adverbial function
收藏资源简介:
该数据集是由巴黎东大学研究团队构建的法语多词表达式副词功能标注语料库,旨在支持信息检索与抽取以及句法分析研究。语料库包含两个来源:法国国民议会2006年10月3-4日会议完整记录(AS)和儒勒·凡尔纳小说《八十天环游地球》(JV),总计约168,846个词符和37,856个词型,标注了4,247个多词副词实例。数据集通过结合6,800条目的句法语义词典与有限状态转换器自动标注,并经过人工校验流程构建而成,主要应用于自然语言处理领域,解决多词副词识别、介词附着消歧等关键问题,为法语计算语言学资源提供重要补充。
This dataset is a French multi-word expression adverbial function annotated corpus developed by the research team of Université Paris-Est, aiming to support research in information retrieval and extraction as well as syntactic parsing. The corpus draws from two sources: the complete transcripts of the French National Assembly sessions held on October 3–4, 2006 (abbreviated as AS), and Jules Verne's novel *Around the World in Eighty Days* (abbreviated as JV). In total, it contains approximately 168,846 tokens and 37,856 types, with 4,247 multi-word adverb instances annotated. Constructed via automatic annotation using a syntactic-semantic lexicon of 6,800 entries combined with finite-state transducers, the dataset was further refined through a manual validation workflow. It is primarily utilized in the field of natural language processing, tackling core tasks such as multi-word adverb recognition and preposition attachment disambiguation, and constitutes a valuable supplementary resource for French computational linguistics.

- 该法语语料库首次构建并公开发布,旨在标注具有副词功能的多词表达,为自然语言处理研究提供专门的标注资源。
- 语料库被应用于多词表达识别与解析任务,验证了其在法语副词性短语自动抽取中的有效性,成为相关领域的重要基准数据集。
- 研究者基于该语料库开展了跨语言对比分析,探讨法语副词性多词表达在机器翻译中的处理策略,拓展了其应用范围。
- 语料库被纳入法语语言资源共享平台,支持更广泛的学术研究,并推动了多词表达标注规范的标准化进程。
- 1A French corpus annotated for multiword expressions with adverbial functionUniversité Paris Nanterre, Université Paris-Saclay · 2018年
- 2Multiword Expressions in Natural Language Processing: A SurveyUniversity of Stuttgart, University of Tübingen · 2021年
- 3Annotation of Multiword Expressions in French: A Comparative StudyUniversité Paris Nanterre, CNRS · 2020年
- 4French Adverbial Multiword Expressions: A Corpus-Based AnalysisUniversité Paris-Saclay, Université Paris Nanterre · 2019年
- 5MWE-FR: A Large-Scale French Multiword Expression LexiconUniversité de Lorraine, CNRS · 2022年



