TagClusters: Enhancing Semantic Understanding of Collaborative Tags

TagClusters: Enhancing Semantic Understanding of Collaborative Tags

Ya-Xi Chen (Media Informatics, University of Munich, Germany), Rodrigo Santamaría (Department of Informatics and Automatics, University of Salamanca, Spain), Andreas Butz (Media Informatics, University of Munich, Germany) and Roberto Therón (Department of Informatics and Automatics, University of Salamanca, Spain)
DOI: 10.4018/978-1-4666-0285-4.ch007
OnDemand PDF Download:
No Current Special Offers


Many online communities use TagClouds, an aesthetic and easy to understand visualization, to represent popular tags collaboratively generated by their users. However, due to the free nature of tagging, such collaborative tags have linguistic problems and limitations, such as high semantic density. Moreover, the alphabetical order of TagClouds poorly supports a hierarchical exploration among tags. This paper presents an exploration to support semantic understanding of collaborative tags beyond TagClouds. Based on the results of the authors’ survey of practical usages of collaborative tags, they developed a visualization named TagClusters, in which tags are clustered into different groups, with font size representing tag popularity and the spatial distance indicating the semantic similarity between tags. The subgroups in each group and the overlap between groups are highlighted, illustrating the underlying hierarchical structure and semantic relations between groups. The authors conducted a comparative evaluation with TagClouds and TagClusters based on the same tag set. The results confirmed the advantage of TagClusters in facilitating browsing, comparing and comprehending semantic relations between tags.
Chapter Preview


TagClouds is a visualization of popular tags (see Figure 1). In this visualization, tags are normally ordered alphabetically, with font size representing their popularity. Additional information, such as recency can be illustrated with color lightness. Comparing with tags generated by single users, these popular tags in TagClouds have a higher degree of generality and accuracy. Such visualization facilitates quickly foraging an overall impression of the most popular items, and thus conveys the general interests among a large audience (Viégas & Wattenberg, 2008), (Hearst 2008), (Hearst & Rosner, 2008). TagClouds can be also used for keyword-based search by selecting one or multiple tags as input.

Figure 1.

TagClouds in Last.fm


Limitations Of Tagclouds

In most online communities, users are allowed to generate tags freely, without any restriction or quality control. This low usage barrier has attracted hundreds of millions users. However, due to the free nature of tagging, some problems are inevitable with these collaborative tags (Li, Bao, Yu, Fei & Su, 2007; Hassan & Herrero, 2006).

Linguistic Problems

Nielsen (2007) and Begelman, Keller and Smadja (2006) discovered educational and cultural background influence people’s understanding of tags, which is one of the reasons for tag inconsistency among different users. With no input restrictions, two general problems are hardly to avoid from the users’ perspective (Wu, Zhang, & Yu, 2006): Synonymy, which is also termed as “inter-indexer inconsistency” by Nielsen (2007), appears when different terms are used to describe the same item. A term with several different meanings brings ambiguity (Mathes, 2008), which may reduce the precision of the retrieval results.

Complete Chapter List

Search this Book: