跨语言信息检索中的查询扩展

Query Expansion in Cross-Language Information Retrieval

提出了把语义词典和马尔可夫随机域的潜在语义扩展相结合的新方法,充分利用了现有词典资源和语料库资源,又克服了单一使用词典方法的局限性和单一使用语义扩展的不确定性,扩展出查询词的同义近义词,上位词和潜在语义相关词,实验表明该方法能很好的解决跨语言信息检索中翻译项的歧义问题。

this paper proposes a new method which integrates with semantic dictionary and Latent Concept Expansion Using Markov Random Fields .The method can make the best of the dictionary and corpus ,overcome the shortcoming of using dictionary only and using Concept Expansion only .Expand original query with related , latent concept words and antonyms completely .The experimental result show that the new method can resolve the problem of different meanings in Cross-Language Information Retrieval..