Mining Clickthrough Data for Collaborative Web Search

• Jian-Tao Sun, Microsoft Research Asia, P.R. China
• Xuanhui Wang, Department of Computer Science, University of Illinois at Urbana-Champaign, USA
• Dou Shen, Department of Computer Science, Hong Kong University of Science and Technology, P.R. China
• Hua-Jun Zeng, Microsoft Research Asia, P.R. China
• Zheng Chen, Microsoft Research Asia, P.R. China

Full text:

Track: Posters

Web search can be regarded as a social behavior as users all over the world seek information from the Web by search engines. The purpose of this paper is to investigate the group behavior patterns of search activities based on the clickthrough data, to boost search performance. We proposed a Collaborative Web Search (CWS) framework based on the probabilistic modeling of the heterogeneous clickthrough objects, including users, queries and Web pages. The CWS framework consists of two steps: first, a cube-clustering approach is put forward to estimate the semantic cluster structures of the clickthrough data objects; next, Web search activities are conducted by leveraging the probabilistic relations among the cluster structures. Experiments on a real-world clickthrough data set validate the effectiveness of our CWS approach.

Other items being presented by these speakers

• A Comparison of Implicit and Explicit Links for Web Page Classification (Data Mining Track)
• CWS: A Comparative Web Search System (Search Track)