TY - GEN
T1 - How useful are tags? - An empirical analysis of collaborative tagging for web page recommendation
AU - Zeng, Daniel
AU - Li, Huiqian
PY - 2008/7/1
Y1 - 2008/7/1
N2 - As a representative Web 2.0 application, collaborative tagging has been widely adopted and inspires significant interest from academies. Roughly, two lines of research have been pursued: (a) studying the structure of tags, and (b) using tag to promote Web search. However, both of them remain preliminary. Research reported in this paper is aimed at addressing some of these research gaps. First, we apply complex network theory to analyze various structural properties of collaborative tagging activities to gain a detailed understanding of user tagging behavior and also try to capture the mechanism that can help explain such tagging behavior. Second, we conduct a preliminary computational study to utilize tagging information to help improve the quality of Web page recommendation. The results indicate that under the user-based recommendation framework, tags can be fruitfully exploited as they facilitate better user similarity calculation and help reduce sparsity related to past user-Web page interactions.
AB - As a representative Web 2.0 application, collaborative tagging has been widely adopted and inspires significant interest from academies. Roughly, two lines of research have been pursued: (a) studying the structure of tags, and (b) using tag to promote Web search. However, both of them remain preliminary. Research reported in this paper is aimed at addressing some of these research gaps. First, we apply complex network theory to analyze various structural properties of collaborative tagging activities to gain a detailed understanding of user tagging behavior and also try to capture the mechanism that can help explain such tagging behavior. Second, we conduct a preliminary computational study to utilize tagging information to help improve the quality of Web page recommendation. The results indicate that under the user-based recommendation framework, tags can be fruitfully exploited as they facilitate better user similarity calculation and help reduce sparsity related to past user-Web page interactions.
UR - http://www.scopus.com/inward/record.url?scp=45849127609&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=45849127609&partnerID=8YFLogxK
U2 - 10.1007/978-3-540-69304-8_32
DO - 10.1007/978-3-540-69304-8_32
M3 - Conference contribution
AN - SCOPUS:45849127609
SN - 3540691367
SN - 9783540691365
T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
SP - 320
EP - 330
BT - Intelligence and Security Informatics - IEEE ISI 2008 International Workshops
T2 - IEEE International Conference on Intelligence and Security Informatics, ISI 2008 Workshops: PAISI, PACCF, and SOCO 2008
Y2 - 17 June 2008 through 17 June 2008
ER -