跨语言信息检索查询翻译消歧方法与技术研究

A Study on Disambiguation Methods and Technologies for Query Translation of Cross-language Information Retrieval

文章从翻译资源消歧的角度,探讨跨语言信息检索的查询翻译的消歧方法与技术。其中,方法包括基于词典、语料库、泛众协作知识源以及本体知识库的方法;具体技术涉及词性标注与词法分析、专有名词的音译与双向翻译技术、互信息模型的应用、利用虚拟语境消除多义性、查询扩展技术、查询分类以及相关反馈等技术。并进一步讨论了翻译消歧中多种方法技术的综合利用,以期为我国此领域的研究提供借鉴。

From the perspective of translation resources disambiguation, this article discusses methods and technologies for query translation disambiguation of cross-language information retrieval. Among them, the method includes some based on dictionary and corpus, the coordination tacit knowledge and ontology knowledge base; specific techniques involves part-of-speech tagging and lexical analysis, proper nouns transliteration, and two-way translation technology, the application of mutual information model, using the virtual context to eliminate ambiguity, query expansion, query classification and relevant feedback technology. And comprehensive utilization of a variety of methods and techniques are further discussed, in order to orovide reference to the research in our country in this field.