An evaluation of structure-preserving and query-biased summaries in web search tasks

Automatic summarization has started to receive increasing attention in recent years due to the increased amount of information available in electronic form. Especially, summarization techniques can be very useful in improving the effectiveness of information retrieval on the World Wide Web. However, currently available major search engines such as Google show only a limited capability for summarization. We believe that text summarization with more sophisticated techniques can significantly improve the search experience of users. As a novel approach, we propose a query-biased summarization system which incorporates the structure of documents into the summaries. The system also makes use of natural language processing techniques in the summarization process. The effectiveness of the proposed system has been tested on a task-based evaluation.