Tagging semantics: investigations with wordnet

ABSTRACT

The content of a tag sequence references both a user's concepts and the user's conceptualization of an information object. The tagging history of 823 users of the Delicious social tagging service is analyzed using WordNet. Three semantic measures of the tagging content are developed: the level of category references, the changes in category level for each noun as the tagging sequence unfolds, and the scope of concept coverage as the compactness of the WordNet subgraph for the noun senses. Observed patterns of concept reference as a function of sequence position hint at dynamic properties of the tag production process by marking a trace of cognitive activity. If tagging is object categorization, these measures provide a view of the personal categorization behavior of non-professionals and illuminate biases in the production of 'folksonomies' due to tag production processes.

INDEX TERMS

Primary Classification:
  H. Information Systems
  H.3 INFORMATION STORAGE AND RETRIEVAL
      H.3.1 Content Analysis and Indexing
          Subjects: Indexing methods

General Terms:
Human Factors, Theory

Keywords:
categorization, folksonomy, semantics, tagging