Generative Language Model: From Text Retrieval to Named Entity Retrieval
Language modeling techniques have been widely used in the domain of text retrieval. This paper presents a generative model that makes use of language modeling techniques for named entity retrieval. Such model is based on text retrieval and extends it to search named entities. In the model, a language model for each named entity is estimated with a set of documents using language modeling techniques. Then the generative probability of a query being caused by a document is calculated, as well as the generative probability of a document being caused by an entity. Such a set of twostage generations give the overall estimation of how the named entity is relevant to the query. Our evaluation on an expert finding task shows that this general framework is comparable to some sophisticated approaches.
Resource management data mining data association
Zhao Ru Jun Guo Weiran Xu
School of Information Engineering Beijing University of Posts and Telecommunications Beijing, China, 100876
国际会议
The International Colloquium on Onformation Fusion 2007(2007年国际信息融合研讨会)
西安
英文
490-494
2007-08-22（万方平台首次上网日期，不代表论文的发表时间）