Classification-enhanced ranking
Authors:
Open Access
- 26 April 2010
- conference paper
- Published by Association for Computing Machinery (ACM)
- p. 111-120
Abstract
Many have speculated that classifying web pages can im- prove a search engine's ranking of results. Intuitively re- sults should be more relevant when they match the class of a query. We present a simple framework for classication- enhanced ranking that uses clicks in combination with the classication of web pages to derive a class distribution for the query. We then go on to dene a variety of features that capture the match between the class distributions of a web page and a query, the ambiguity of a query, and the coverage of a retrieved result relative to a query's set of classes. Ex- perimental results demonstrate that a ranker learned with these features signicantly improves ranking over a compet- itive baseline. Furthermore, our methodology is agnostic with respect to the classication space and can be used to derive query classes for a variety of dierent taxonomies.Keywords
This publication has 24 references indexed in Scilit:
- LETOR: A benchmark collection for research on learning to rank for information retrievalInformation Retrieval Journal, 2010
- Adapting boosting for information retrieval measuresInformation Retrieval Journal, 2009
- Classifying search queries using the Web as a source of knowledgeACM Transactions on the Web, 2009
- Automatic classification of Web queries using very large unlabeled query logsACM Transactions on Information Systems, 2007
- KDD CUP-2005 reportACM SIGKDD Explorations Newsletter, 2005
- Classifying search engine queries using the web as background knowledgeACM SIGKDD Explorations Newsletter, 2005
- The Ferrety algorithm for the KDD Cup 2005 problemACM SIGKDD Explorations Newsletter, 2005
- Q 2 C@USTACM SIGKDD Explorations Newsletter, 2005
- Learning to rank using gradient descent2005
- Agglomerative clustering of a search engine query log2000