Abstract: In an IR model, “user”, “query” and “result” are three important components. Traditionally, a query is considered to be independent of the user. IR systems search documents without considering who issues the query and why the query is asked. However, those factors can affect the user’s satisfaction about the result. The information about the user, such as the genre preference of the user, occupation of the user, location of the user, which are normally called personalized information, indicate users’ preferences to the retrieval result. We demonstrate the effectiveness of a dual indexing technique and a feedback method on the HARD 2005 data set. We also propose a non-content based method to measure user’s familiarity to a query. A similarity model is built to utilize the familiarity information and to improve the overall performance.
Publication Year: 2005
Publication Date: 2005-01-01
Language: en
Type: article
Access and Citation
Cited By Count: 2
AI Researcher Chatbot
Get quick answers to your questions about the article from our AI researcher chatbot