Query Suggestion

Abstract

Query suggestion is one of the few fundamental problems in Web search. It assists users to refine queries in order to satisfy their information needs. Many query suggestion techniques have been proposed in the past decades. The mainstream idea is to leverage query logs which contain the search behaviors of users to generate useful query suggestions. In this chapter, we introduce several log-based query suggestion techniques. These methods fall into four categories: (1) query co-occurrence; (2) query-URL bipartite graph; (3) query transition graph; and (4) short-term search context. We also briefly discuss other related work in this field and point out several future directions.