Tag co-occurrence analysis using the association data mining rule

  • Authors:
  • Kyunghye Yoon

  • Affiliations:
  • State University of New York at Oswego, Oswego, NY

  • Venue:
  • Proceedings of the 2012 iConference
  • Year:
  • 2012

Quantified Score

Hi-index 0.06

Visualization

Abstract

This paper presents a study in progress on tag co-occurrence employing data mining algorithms. Based on the assumption that the tag terms occurring together are closely related with each other, the study intended to investigate the relationship of tag terms that appear together in a tag set. The association data mining rule was used to find the tag pairs that occur frequently together to identify those of relatively strong association. Analysis was followed to look at the semantic relations of the terms in the selected tag pairs. Preliminary data analysis found that the two tag terms appearing together with stronger association tended to be related syntagmatically with each other (i.e., they were from different concepts of terms that related to the context of use) rather than conceptually similar terms in taxonomic relations.