Context-Aware Basic Level Concepts Detection in Folksonomies
This paper deals with the problem of exploring implicit se mantics in folksonomies. In folksonomies, users create and manage tags to annotate web resources. The collection of usercreated tags in folk sonomies is a potential semantics source. Much research has been done to extract concepts, and even concepts hierarchy (ontology), which is the important component for knowledge representation (e.g. in semantic web and agent communication), from folksonomies. However, there has been no metric for discovering human acceptable and agreeable concepts, and thus many concepts extracted from folksonomies by existing approaches are not natural for human use. In cognitive psychology, there is a fam ily of concepts named basic level concepts which are frequently used by people in daily life, and most human knowledge is organized by basic level concepts. Thus, extracting basic level concepts from folksonomies is more meaningful for categorizing and organizing web resources than extracting concepts in other granularity. In addition, context plays an important role in basic level concepts detection, as the basic level con cepts in the same domain become different in different contexts. In this paper, we propose a mcthod to detect basic level concepts in different contexts from folksonomies. Using Open Directory Project (ODP) as the bcnchmark, we demonstratc the existence of context effect and the effectiveness of our method.
Wen-hao Chen Yi Cai Ho-fung Leung Qing Li
Department of Computer Science and Engineering The Chinese University of Hong Kong, Hong Kong, China Department of Compnter Science City University of Hong Kong, Hong Kong, China
国际会议
11th International Conference,WAIM 2010(第十一届网络时代管理国际会议)
九寨沟
英文
632-643
2010-07-14（万方平台首次上网日期，不代表论文的发表时间）