Thefolksonomytagcloudwhenisituseful.pptVIP

The folksonomy tag cloud: when is it useful? James Sinclair and Michael Cardew-Hall Department of Engineering, The Australian National University, Canberra, Australia Journal of Information Science (JIS), 2007 Outline Introduction Experimental Method Results Conclusion Introduction Tagging Services Tagging in the enterprise Tag Clouds Tagging Services Tagging services allow participants to associate freely determined keywords (called ‘tags’) with a particular resource. There is little incentive for authors to put time and effort into providing quality metadata unless it is for the purposes of self-promotion. Thus, author generated metadata is perceived as inaccurate at best, and deliberately misleading at worst. [Thomas and Griffin, 1999] Tagging in the enterprise The sharing of key resources and the generation of emergent vocabulary helps decrease the cognitive distance between people working in the organization. Tag Clouds User interface element commonly associated with folksonomy datasets. Tomas Vader Wal --- the man credited with coining the term ‘folksonomy’ --- described tag clouds as being ‘cute but offering little value’ Do tag clouds provide value to people seeking information from a folksonomy dataset? Experimental Method Participant: provided with 2 options to perform information-seeking task Tag cloud Search box Tag around 10 articles each 2 session: tag and query Hypothesis: if people found no value in a tag cloud for information seeking, then they would primarily use the search box to seek information Participant Demography Tagging Interface: Session One Tag Size Tag cloud displayed top 70 most-used tags Folksonomy frequency tends to follow a power law distribution Question List Search Interface: Session Two Tag Cloud Query Exit Survey Dataset Characteristics Number of participants relatively small Participant not select articles to tag No feedback aspect to tagging. Participant not able to browse other’s tags Result Assume last query made before a