Parallel Meaning Bank
收藏资源简介:
Parallel Meaning Bank(PMB)是由格罗宁根大学创建的多语言翻译数据集,包含超过1100万字的英语、德语、意大利语和荷兰语文本,总计285,154个文档。该数据集通过跨语言投影方法,利用英语的自动语义标注映射到其他语言,确保翻译的语义保持性。数据集内容丰富,涵盖多种文本类型和语言对,旨在支持跨语言语义分析和语义解析器的开发,以及研究翻译过程中的语义变化。创建过程涉及文本分割、句法分析、语义标注等多个步骤,使用统计模型进行半监督学习。PMB的应用领域广泛,包括语言学研究、机器翻译和自然语言处理等,旨在解决多语言环境下的语义理解和翻译问题。
Parallel Meaning Bank (PMB) is a multilingual translation dataset created by the University of Groningen. It contains over 11 million words of English, German, Italian and Dutch texts, totaling 285,154 documents. It adopts a cross-language projection approach that maps automatic semantic annotations derived from English to other languages, ensuring the semantic fidelity of the translated content. Featuring rich content covering diverse text types and language pairs, the dataset aims to support cross-language semantic analysis, the development of semantic parsers, and research on semantic changes during the translation process. Its creation encompasses multiple stages including text segmentation, syntactic parsing, semantic annotation and more, utilizing statistical models for semi-supervised learning. PMB has wide-ranging applications in fields such as linguistic research, machine translation and natural language processing, with the core goal of addressing semantic understanding and translation challenges in multilingual scenarios.
- 1The Parallel Meaning Bank: Towards a Multilingual Corpus of Translations Annotated with Compositional Meaning Representations格罗宁根大学 · 2017年



