Using Association Rules for Expanding Search Engine Recommendation Keywords in English and Chinese Queries

Y.P. Huang, C.-A. Tsai (Taiwan), and F.E. Sandnes (Norway)


Information retrieval, query modification, web search, personalization service


In this paper, a new search algorithm is proposed which exploits association rules to enhance users’ search engine experience. After users submit query keywords, more detailed expansion words are determined using Google and Yahoo and returned to the users. In addition to handling simple search terms, the proposed system is also capable of processing bilingual phrase queries in English and Traditional Chinese. Each word is treated as an item. Then, by computing the confidence value of one term and its following terms, we can determine whether they can be concatenated into a phrase or not. The phrase with a confidence value greater than a given threshold is treated as a large item set which represents the consecutive appearance in an article. Personalization is achieved by using the grey relational method to rank the users' preferences according to past behavior recorded in log files. Association rules are extracted to construct a recommendation mechanism. Detailed simulation results are given to illustrate the feasibility and applicability of the presented work.

Important Links:

Go Back