Development of tag ranking-based image search system

Abstract

The existing tag-based systems offer search results with low accuracy, as they provide search results through a single tag matching by using content-tagged tags. Because users do not consider correlations and priorities between tags upon content tagging, the content and relevant information that the tags have are not efficiently offered. In this context, this study suggests a cluster-based tag ranking and search system using tag similarity to solve the problems mentioned above. Through such a system, the existing single tag matching problem can be solved. © 2014 SERSC.