The development and validation of a one-bit comparison for evaluating the maturity of tag distributions in a Web 2.0 environment

• Published in: Journal of the American Society for Information Science and Technology Vol. 67; no. 6; pp. 1430 - 1445
• Main Authors: Tang, Kuo-Hao, Tsai, Li-Chen, Hwang, Sheue-Ling
• Format: Journal Article
• Language: English
• Published: Hoboken Blackwell Publishing Ltd 01.06.2016; Wiley Periodicals Inc
• Subjects: Comparative analysis; Cues; Design analysis; domain analysis; Entropy; Entropy (Information theory); entropy (information); Experts versus novices; Learning; Organizations; Recommender systems; Scientific Concepts; Social Influences; Studies; Subject specialists; Tagging; Tags; text mining; Web 2.0; Webs
• ISSN: 2330-1635; 2330-1643
• DOI: 10.1002/asi.23454

Tags generated by domain experts reaching a consensus under social influence reflect the core concepts of the tagged resource. Such tags can act as navigational cues that enable users to discover meaningful and relevant information in a Web 2.0 environment. This is particularly critical for nonexperts for understanding formal academic or scientific resources, also known as hard content. The goal of this study was to develop a novel one‐bit comparison (OBC) metric and to assess in what circumstances a set of tags describing a hard‐content resource is mature and representative. We compared OBC with the conventional Shannon entropy approach to determine performance when distinguishing tags generated by domain experts and nonexperts in the early and later stages under social influence. The results indicated that OBC can accurately distinguish mature tags generated by a strong expert consensus from other tags, and outperform Shannon entropy. The findings support tag‐based learning, and provide insights and tools for the design of applications involving tags, such as tag recommendation and tag‐based organization.