JRC's Participation in the Guided Summarization Task at TAC 2010

In this paper we describe our participation in the Guided Summarization Task at the Text Analysis Conference 2010 (TAC’10). The goal of the task was to encourage a deeper semantic analysis of the source documents instead of relying only on document word frequencies to select important concepts. We used the output of our event extraction system and automatic learning of semantically-related terms to capture the required aspects of each particular article category. We submitted two runs: the first uses information extraction tools in combination with co-occurrence of features, the second uses only co-occurrence information. In the following sections we describe our runs and discuss the results attained.