Machine learning in automated text categorization
ACM Computing Surveys (CSUR)
Modern Information Retrieval
Enriching web taxonomies through subject categorization of query terms from search engine logs
Decision Support Systems - Web retrieval and mining
Clustering documents in a web directory
WIDM '03 Proceedings of the 5th ACM international workshop on Web information and data management
Clustering documents into a web directory for bootstrapping a supervised classification
Data & Knowledge Engineering - Special issue: WIDM 2003
Reducing human interactions in Web directory searches
ACM Transactions on Information Systems (TOIS)
International Journal of Human-Computer Studies
Web page classification: Features and algorithms
ACM Computing Surveys (CSUR)
Learning to integrate web taxonomies
Web Semantics: Science, Services and Agents on the World Wide Web
From web directories to ontologies: natural language processing challenges
ISWC'07/ASWC'07 Proceedings of the 6th international The semantic web and 2nd Asian conference on Asian semantic web conference
Classifying web data in directory structures
APWeb'06 Proceedings of the 8th Asia-Pacific Web conference on Frontiers of WWW Research and Development
| Hi-index | 0.00 |
Web directories allow Web users to browse a hierarchy of categories, under which di-fferent types of resources are classified. We study the problem of maintaining a Webdirectory, that is, the problem of continually discovering and ranking resources that arerelevant to the categories of the directory. We propose an unsupervised computationalmethod that conducts the maintenance of the directory by analyses of user browsingdata. The method is based on the extraction and classification of user sessions (se-quences of resources selected by users) into the categories of the directory. In addition,we show that the directory maintenance method can be slightly modified to find queriesthat are useful to find relevant resources allowing users to switch from directory browsingto query formulation. Experimental results allow for affirmation that the proposed me-thods are effective, that they attain identification of new pages in each category and alsorecommend related queries with high precision, without needing labeled data to conducttraditional web page and query classification tasks.