Conference Paper

Query Expansion Based on a Personalized Web Search Model

Chongqing Univ., Chongqing;
DOI: 10.1109/SKG.2007.83 Conference: Semantics, Knowledge and Grid, Third International Conference on
Source: IEEE Xplore

ABSTRACT A novel query expansion algorithm is proposed in this paper. It is based on a model of personalized web search system. The new system, as a middleware between a user and a Web search engine, is set up on the client machine. It can learn a user's preference implicitly and then generate the user profile automatically. When the user inputs query keywords, more personalized expansion words are generated by the proposed algorithm, and then these words together with the query keywords are submitted to a popular search engine such as Baidu or Google. These expansion words can help a search engine retrieval information for a user according to his/her implicit search intentions. The new Web search model can make a common search engine personalized, that is, throughout personalized query expansion the search engine can return different search results to different users who input the same keywords. The experimental results show the effect and applicability of the presented work for personalized information service of a search engine.

  • [Show abstract] [Hide abstract]
    ABSTRACT: Traditional search systems provide users with a starting point for their information search. Information Retrieval (IR) Systems present only initial results list. The relevance measure of documents and understanding user's queries are main issues in design of IR Systems. This involves improvements at query level and result display level. Personalized retrieval widens the notion of information need to comprise implicit user needs, not directly conveyed by the user in terms of explicit information requests. This can be achieved by expanding the user query and processing the results according to the user needs. Individual pages are retrieved by the traditional IR systems even though the information is spread across multiple pages. Instead composed pages are generated which contains all the query words. Ranking of retrieved pages can be improved by providing composed pages for the given query. As Agents can provide autonomous functioning, they can be used in the design of query expansion, searching and ranking of documents. This paper proposes a multi agent-based intelligent retrieval framework for query expansion and composing of pages.
    Technology Enhanced Education (ICTEE), 2012 IEEE International Conference on; 01/2012
  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: Area of mobile web search personalization is gaining demand due to relevancy of results. Generally user relevance is captured in the form of profiles. Profile generated implicitly according to user search history or explicitly using appropriate interface. We propose personalized mobile search engine that take best features of both implicit & explicit user profiles for personalization. Implicit user profile is ontology-based & multifaceted. Concepts are divided into content concepts & location concepts to understand importance of location in the personalization. GPS locations are also used for better personalization. In this paper, explicit user profiles created, deleted & edited through user interface. System evaluates relevancy of search results using Normalized Kendall Tau Distance, precision & recall. Experimental results show that effectiveness of personalization has improved.
    International Conference on Industrial Electronics and Computer Engineering at New Delhi, New Delhi, India; 08/2014
  • [Show abstract] [Hide abstract]
    ABSTRACT: Web Search Engines have become a resource pool for the people to gain information. They always aim at achieving the goal of delivering relevant information to the users. Due to the exponential growth of information and imprecise queries, Search engines could not meet the user's information requirements. Users have to reframe their queries until they get their desired information. Also results are not listed based on the users search context. Personalization is re-ordering of the search results that satisfies the needs of the user. This paper presents an approach to capture the user's preferences in order to provide query suggestions. Web search personalization using collaborative filtering adapts a generic search engine for the needs of a community of users. The main objective of this work is to improve the retrieval of information by expanding the user query and to rank the result list based on the users domain of interest.