An approach to sentence-selection-based text summarization

Abstract

We introduced a newly developed text summarization system. It supports both Chinese and English, while this paper focuses on Chinese processing. We apply 6 word level features and 3 sentence level features to weigh each word and sentence. We also describe two new techniques, one is for processing the topic sensitive word feature and another is for processing the sentence length feature. Primary subjective evaluation shows that these approaches are effective and efficient, and performance of the system is promising.