Tagging behaviour with support from controlled vocabulary

Social tagging is a way for users to provide user-oriented access to information resources on the Web. In information retrieval, tagging may be complementary to traditional indexing methods. However, many of the existing social tagging applications have not been designed with information retrieval in mind. In the EnTag project (Enhancing Tagging for Discovery) we addressed the problem of lack of specificity and novelty of tags. The project investigated ways of enhancing social tagging via controlled vocabularies, with a view to improving the suitability for information discovery and retrieval. This paper investigates how knowledge structures from a controlled vocabulary affect end-users’ assignment of tags. The study is a comparison of tags assigned using a traditional tagging system that provides suggestions from two tag clouds (all users’ tags and each user’s tags) and an experimental, enhanced tagging system that additionally offers suggestions from the Dewey Decimal Classification system (DDC). We specifically address the following research question: Is it possible to identify differences between tags regarding specificity, exhaustivity, and overlap with title terms and assigned controlled and uncontrolled keywords when using only social tagging versus when using social tagging in combination with suggestions from a controlled vocabulary