Social Search and Task-Related Relevance Dimensions in Microblogging Sites

Social media, and in particular microblogging sites, allow users to post multiple kinds of content for different purposes. Content may be purely conversational, or news-related, or event-related. To find information relevant to users in this heterogeneous mass of content, it would be important to consider the task for which search is carried out, and the most suitable relevance dimensions. In the last years, despite the social search problem has been increasingly investigated, this aspect has not been sufficiently analyzed. For this reason, in this paper, we focus on different search tasks in the microblog search context, and we identify some related relevance dimensions. We also report some experiments we have made to verify the impact of the identified relevance dimensions on the system effectiveness, with respect to the considered search tasks.