抽象的

Mining User Profile Using Clustering From Search Engine Logs

DR.A.MUTHU KUMARAVEL

Fundamental component of any personalization application is user profiling. The existing user profiling strategies are based on users interest (i.e. positive preferences).The main focus is on search engine personalization and to develop several concept-based user profiling methods. Concept-based user profiling methods deals with both positive and negative preferences. This user profiles can be integrated into the ranking algorithm of a search engine so that search result can be ranked according to individual users interest. The RSCF makes a search of data containing the item in the search results, the required data is been clicked by the user and this clicked data is given as the input and generates the rankers as the output.. The negative preference increases the separation between the similar and dissimilar queries. This separation provides a clear threshold for agglomerative clustering algorithm and improves the overall quality.

免责声明: 此摘要通过人工智能工具翻译,尚未经过审核或验证