It is obvious that web search queries given by the user are always short and ambiguous. Mostly the shorter length queries do not satisfy the users real information need and may not produce the results properly. Query Recommendation is a technique based on the real intent of the user and to provide the alternate queries to frame the queries in the future. The proposed work recommends the queries for four types of users in three ways (1) Favourite queries of the user are identified and they are recommended. (2) Users who have similar interest are clustered; the recommendation is given from the access logs of similar users. (3) Similar queries are clustered; the favourite query of the cluster is identified and it is recommended. The proposed work also ranks the recommended queries based on the preference and access time of the query. The proposed strategies are experimentally evaluated using real time search engine query log.
展开▼
