Name
Abstract | ||
|---|---|---|
An approach to clustering short text snippets is proposed, which can be used to cluster search results into a few relevant groups to help users quickly locate their interesting groups of results. Specifically, the collection of search result snippets is regarded as a similarity graph implicitly, in which each snippet is a vertex and each edge between the vertices is weighted by the similarity between the corresponding snippets. TermCut , the proposed clustering algorithm, is then applied to recursively bisect the similarity graph by selecting the current core term such that one cluster contains the term and the other does not. Experimental results show that the proposed algorithm improves the KMeans algorithm by about 0.3 on FScore criterion. |
Year | DOI | Venue |
|---|---|---|
2009 | 10.1007/978-3-642-00672-2_55 | APWeb/WAIM |
Keywords | Field | DocType |
fscore criterion,search results,short text clustering,current core term,similarity graph,corresponding snippet,search result snippet,cluster search result,kmeans algorithm,proposed clustering algorithm,proposed algorithm,text clustering,graph partitioning | Canopy clustering algorithm,Fuzzy clustering,Data mining,CURE data clustering algorithm,Data stream clustering,Pattern recognition,Correlation clustering,Computer science,Constrained clustering,Artificial intelligence,Cluster analysis,Single-linkage clustering | Conference |
Volume | ISSN | Citations |
5446 | 0302-9743 | 3 |
PageRank | References | Authors |
0.42 | 6 | 5 |
Name | Order | Citations | PageRank |
|---|---|---|---|
| Xingliang Ni | 1 | 75 | 3.71 |
| Zhi Lu | 2 | 257 | 11.74 |
| Xiaojun Quan | 3 | 260 | 20.64 |
| Liu Wenyin | 4 | 2531 | 215.13 |
| Bei Hua | 5 | 146 | 17.17 |