Detecting Hot Events from Web Search Logs

摘要：

Detecting events from web resources is a challenging task, attracting many attentions in recent years. Web search log is an important data source for event detection because the information it contains reflects users activities and interestingness to various real world events. There are three major issues for event detection from web search logs: effectiveness, efficiency and the organization of detected events. In this paper, we develop a novel Topic and Event Detection method, TED, to address these issues. We first divide the whole data into topics for efficiency consideration, and then incorporate link information, temporal information and query content to ensure the quality of detected events. Finally, events detected are organized through the proposed interestingness measure as well as topics they belong to. Experiments are conducted on a commercial search engine log. The results demonstrate that our method can effectively and efficiently detect hot events and give a meaningful organization of them.

关键词： Event detection click-through data efficient algorithm

作者: Yingqin Gu Jianwei Cui Hongyan Liu Xuan Jiang Jun He Xiaoyong Du Zhixu Li

作者单位: Key Labs of Data Engineering and Knowledge Engineering, Ministry of Education, China School of Infor Department of Management Science and Engineering, Tsinghua University, Beijing 100084, China

会议类型: 国际会议

会议名称: 11th International Conference,WAIM 2010(第十一届网络时代管理国际会议)

会议地点: 九寨沟

会议语种:英文

页码: 417-428

在线出版日期: 2010-07-14（万方平台首次上网日期，不代表论文的发表时间）