Skip to main content Accessibility help
  • Print publication year: 2018
  • Online publication date: June 2019

3 - Social tags for linked data with Resource Description Framework (RDF)



Social tags are often considered as another type of metadata, since they provide information about the resources for users’ needs (Kim et al., 2010; Passant et al., 2009). In addition, its social aspect allows tags to become a form of collective knowledge. The term folksonomy, a term created by combining ‘folk’ and ‘taxonomy’ to indicate a user-driven collective taxonomy generated from tags (Vander Wal, 2007), also reflects such perspectives.

As simple keywords assigned to a resource, tags have some limitations as resource descriptions. First, simple keyword tags without any context encounter problems with ambiguity in their semantic meanings (Guy and Tonkin, 2006; Passant et al., 2009). A good example of this limitation is homograph terms such as apple, which can mean a fruit or a company. Second, because the tagging systems allow users to input free form text as tags, the tag terms may vary in their forms and selections. This could be related to the use of synonymous terms such as car and automobile, where users can select one or the other for the same concept. Another example could be related to different tag term forms. For instance, user-generated compound tag forms such as semantic web, SemanticWeb, Semweb, semantic-web and so forth, are considered heterogeneous in tagging systems and when computers process tags (Guy and Tonkin, 2006; Passant et al., 2009; Syn, 2014). In addition to these cases, there can be additional problems of tag variations that occur due to human errors such as misspelling. Third, a folksonomy does not have a concrete structure; unlike the hierarchical structure of controlled vocabularies, a folksonomy is considered to be a flat organization (Shirky, 2005). In addition, depending on users’ contributions, a tag set may change over time, making a folksonomy dynamic (Lin et al., 2012; Mika, 2007). Finally, there is an issue with the use of tag sets, as user tags are dependent on tagging systems (Gruber, 2007; Kim et al., 2010; Mika, 2007). Various types of systems and applications offer tagging features, including bookmarking systems, such as CiteULike (, photo-sharing systems, such as Flickr ( and Instagram (, microblogging systems, such as Twitter ( and library online public access catalogue (OPAC) systems, such as BiblioCommons (