期刊名称：Journal of Theoretical and Applied Information Technology
印刷版ISSN：1992-8645
电子版ISSN：1817-3195
出版年度：2014
卷号：59
期号：3
出版社：Journal of Theoretical and Applied
摘要：Keyword-based web search is widely used for locating information on the web. But, web users lack sufficient domain knowledge and find it difficult to organize and formulate input queries which affect search performance. Existing method suggests terms using the statistics in the documents, query logs and external dictionaries. This novel query suggestion method suggests terms related to topics present in the input query and re-rank the retrieved documents. A generative model, Latent Dirichlet Allocation (LDA) is used to learn the topics from the underlying documents. The high probability words in a topic are selected using the Kullback liebler(KL) divergence measure and presented to the user for suggestion, to enrich the user query and to narrow the search. The re-ranking technique of this approach uses the initial retrieval position of the document to re-rank the documents. The suggested queries by the hidden topic approach and by keyword search are analysed.
关键词：Latent Dirichlet Allocation; Kullback-Liebler Divergence; Query Suggestion; Web Search
