LINKING RELATED CONTENT IN WEB ENCYCLOPEDIAS WITH SEARCH QUERY TAG CLOUDS Christoph Trattner * Knowledge Management Institute and Institute for Information Systems and Computer Media Graz University of Technology Inffeldgasse 16c, 8010 Graz, Austria ABSTRACT In this paper we present a novel tool for exploring related resources in Web encyclopedias called QueryCloud. Typically, users come to an encyclopedia from a search engine such as Google, Yahoo! or Bing and upon reading the first page on the site they leave it immediately thereafter. To tackle this problem in other systems such as Web shops, additional browsing tools for easy finding of related content are provided. In order to overcome this issue in the context of Web encyclopedia systems, we introduce a tool called QueryCloud. The tool combines two promising approaches – tag clouds and historic search queries – into a new single one, i.e. each document in the system is enriched with a tag cloud containing collections of related concepts populated from historic search queries. To test the feasibility of the approach, we integrated a prototypical implementation of the tool into a large Web encyclopedia called the Austria-Forum and conducted several experiments on a theoretical and empirical level. As our experiments show, QueryCloud provides a great alternative to traditional forms of tag cloud creation. With several experiments on a theoretical and empirical level we show that tag clouds generated by our system outperform tag clouds that are based on user-tags in terms of linking content and navigability. This work is relevant for researchers interested in the navigability of emergent hypertext structures and for engineers seeking to improve the navigability of large information systems, such as Web encyclopedias. KEYWORDS search query clouds, tag clouds, query tags, tags, linking, web encyclopedia 1. INTRODUCTION Nowadays, content in Web encyclopedias such as Wikipedia is mainly accessed through search engines (Wikimedia 2010). Typically, users with a certain interest in mind go to a search engine such as Google, Yahoo! or Bing, define a search query there and click on a link from the result list from which they are referred to an article within Wikipedia. Upon reading the document they decide to either go back to the search engine to refine their search, or close their browser if they have already found the information they needed. Such a user behavior on encyclopedia sites is traceable through a typical high bounce rate (Alexa 2010, Gleich et al. 2010). Essentially, users do not “really” browse in online encyclopedia systems such as Wikipedia to find further relevant documents (Gleich et al. 2010) - they rather use search engines such as Google, Yahoo! or Bing for that purpose. It is our opinion that Web encyclopedias simply lack usable tools that support users in explorative browsing or searching. For example, in Web based systems such as Web shops, different approaches have been applied to tackle this situation. Amazon for instance offers the user related information through collaborative filtering techniques for each product. Google or Yahoo! apply a similar approach by offering related content (e.g. sponsored links) to the user by taking the users’ search query history into account (Mehta et al. 2007). On the other hand, social tagging systems have emerged as an interesting alternative to find relevant content on the Web (Heymann et al. 2010). These systems apply the concept of social navigation (Millen and * Part of this research was conducted while the author was working as a visiting researcher at the University of Pittsburgh, School of Information Science.