用AI构建阿拉伯语术语库，统一维基百科技术词汇翻译
WikiTermBase: An AI-Augmented Term Base to Standardize Arabic Translation on Wikipedia
- 基于语义与形态分析，整合多源数据构建90万+术语库
- 在阿拉伯语维基百科中标准化英法术语翻译，提升一致性
- 开源工具助力低资源语言技术术语本地化，适合翻译平台开发者
术语库被公认为翻译软件中节省时间并保证一致性的关键组件。尽管自然语言处理和大语言模型近年来取得诸多进展，主流翻译平台仍未充分利用这些技术优化术语库，尤其在支持低资源语言的规模化内容本地化方面仍显不足。阿拉伯世界语言机构自1940年代起就面临科学术语大规模进入阿拉伯语时缺乏统一标准的问题。本文介绍开源工具WikiTermBase，采用系统化方法构建包含90多万条术语的词典数据库，术语来源广泛，基于语义与形态进行收集与映射。该工具已在阿拉伯语维基百科成功应用，实现了对英文与法文术语的标准化翻译。
原文摘要 · Abstract (English)
Term bases are recognized as one of the most effective components of translation software in time saving and consistency. In spite of the many recent advances in natural language processing (NLP) and large language models (LLMs), major translation platforms have yet to take advantage of these tools to improve their term bases and support scalable content for underrepresented languages, which often struggle with localizing technical terminology. Language academies in the Arab World, for example, have struggled since the 1940s to unify the way new scientific terms enter the Arabic language at scale. This abstract introduces an open source tool, WikiTermBase, with a systematic approach for building a lexicographical database with over 900K terms, which were collected and mapped from a multitude of sources on a semantic and morphological basis. The tool was successfully implemented on Arabic Wikipedia to standardize translated English and French terms.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。