Abstract | ||
---|---|---|
Most commercial search engines provide query suggestion in a ranked list for more effective search. However, a ranked list may not be an ideal way to satisfy users' various information demands. In this paper, we propose a novel query suggestion method named CLHQS (Clickthrough-Log based Hierarchical Query Suggestion). It organizes the suggested queries into a well-structured hierarchy. Users can easily generalize, extend or specialize their queries within the hierarchy. The query hierarchy is mined from the clickthrough log data in the following way. First, we generate a candidate set through the query-url graph analysis. Second, the pair-wise relationships are inspected for each pair of candidate queries. Finally, we construct the suggested query hierarchy using these relationships. Experiments on a real-world clickthrough log validate the effectiveness of our proposed CLHQS approach. |
Year | DOI | Venue |
---|---|---|
2009 | 10.1007/978-3-642-01307-2_78 | PAKDD |
Keywords | Field | DocType |
effective search,hierarchical query suggestion,candidate query,query hierarchy,mining clickthrough log,proposed clhqs approach,well-structured hierarchy,suggested query,novel query suggestion method,clickthrough log data,commercial search engine,query suggestion,search engine,satisfiability | Query optimization,Web search query,Data mining,Query language,Query expansion,Ranking,Information retrieval,Computer science,Web query classification,Power graph analysis,Hierarchy | Conference |
Volume | ISSN | Citations |
5476 | 0302-9743 | 0 |
PageRank | References | Authors |
0.34 | 8 | 6 |
Name | Order | Citations | PageRank |
---|---|---|---|
Depin Chen | 1 | 61 | 3.40 |
Ning Liu | 2 | 253 | 15.62 |
Zhijun Yin | 3 | 788 | 37.97 |
Yang Tong | 4 | 0 | 0.34 |
Jun Yan | 5 | 1798 | 85.25 |
Zheng Chen | 6 | 5019 | 256.89 |