Using internal link and social network analysis to support searches in Wikipedia: A model and its evaluation

Wikipedia, the largest multilingual free content online encyclopaedia, is the result of collaboration by countless volunteers around the world. Because of the popularity of such applications, one of the most important issues is how to alleviate information overload on the world wide web and thereby facilitate effective information search activities. Basically, Wikipedia users browse content by following hyperlinks, or they use the search function to look for information. Traditional methods cannot represent all the relationships between hypertexts; thus, users can easily change their search goals unconsciously or get lost when they explore or retrieve information. To address the above problem, we propose a link strength measure that establishes a network by analysing the internal links between articles in Wikipedia. We then apply social network analysis indicators to label the relationships between articles and identify topics to help users find required information. Based on the proposed methods, we developed an application called WikiMap. To assess its effectiveness, we conducted a task-oriented evaluation study with 28 participants who were asked to execute simulated search tasks on two interfaces: the standard Wikipedia interface and the proposed WikiMap interface. The evaluation results show that participants were more successful in executing the assigned tasks on the WikiMap interface, especially unfamiliar decision-making tasks. The exploratory research results have implications for the design of search-aided applications to help users find information in Wikipedia.