WISTUD at TREC 2011: Microblog Track: Exploiting Background Knowledge from DBpedia and News Articles for Search on Twitter

These working notes describe the system developed by the WISTUD team for the Microblog track. We evaluated the suitability of semantic technologies for the search task, in particular, query expansion with named entities that are deduced by means of a topic-based profiling process. The results indicate the feasibility of the approach: for half of the topics, at P@30, our top performing automatic method based on semantic profiling yields better results than the median over all submitted runs.