Browsing and Visualizing Wikipedia through Link Mining and Semantic Relatedness Analysis

With the ubiquity of the Internet and the emergence of Web 2.0 technologies, social web sites (i.e., social networking websites and, micro-blogging services) are providing unprecedented opportunities for creating user-generated content, as well as for promoting communication, collaboration and information-sharing among users. Wikipedia, one of the most famous collaborative projects on the Web, has become an extremely popular reference database for people seeking information or knowledge. However, since the number of articles and the wide variety of topics in Wikipedia is constantly expanding, it is difficult for users to find information efficiently via the hypertext links, i.e., the network of linked documents. To address the problem, we propose a hybrid approach that is based on the theories and techniques of link-based analysis and semantic relatedness analysis. Specifically, we employ a link strength measure to establish a preliminary topic network by analyzing the relationships between articles. We also refine the “Normalized Google Distance” to quantify the strength of the relationship between two articles via key terms. Then, we apply social network analysis indicators to determine the relationships between topics and visualize the analysis results in order to help users browse Wikipedia efficiently. Finally, a topic-based WikiMap is generated based on the proposed hybrid approach. We conducted a user-task oriented evaluation study to confirm that the derived topic-based WikiMap can help users browse topics and execute complicated tasks easily and efficiently.