Answer summarization via term hierarchical structure

In the research area of community-based question and answering (cQA) services such as Yahoo! Answers, most researchers focus on the correctness of answers and pay little attention to the completeness of answers. In this paper, we try to address the answer completeness problem through query-based extractive summarization. Unlike traditional query-based extractive summarization, we extract summary sentences considering not only its relevance to the question but also its relevance to answers and its ability to cover content of other sentences. We model the coverage ability of sentences through the hierarchical structure among terms and assume that terms lying on the upper layers cover more information than terms lying on the lower layers. Also, we propose a novel pyramid-like evaluation method, which considers the completeness, the precision and the conciseness of answers. The experiment results are very promising.