Extracting semantic relations from query logs

• DOI: 10.1145/1281192.1281204
• Publication Date: 2007-12-07T19:19:41Z
• Authors: Ricardo Baeza-Yates; Alessandro Tiberi
• Keywords: Sketch; Semantic space; Spatial Query

In this paper we study a large query log of more than twenty million queries with the goal extracting semantic relations that are implicitly captured in actions users submitting and clicking answers. Previous analyses were mostly done just not followed after them. We first propose novel way to represent vector space based on graph derived from query-click bipartite graph. then analyze produced by our log, showing it is less sparse previous results suggested, almost all measures these graphs follow power laws, shedding some light searching user behavior as well distribution topics people want Web. The representation introduce allows infer interesting relationships between queries. Second, provide an experimental analysis quality relations, most them relevant. Finally sketch application detects multitopical URLs.