Summarizing based on concept counting and hierarchy analysis

We put forward a new summarizing method based on concept counting and hierarchy analysis. By concept extraction and semantic analysis we developed an effective English text summarizing system. This system uses topic concepts to construct a Vector Space Model and partition semantic paragraphs. Combined with readability improvement, the abstract of a text is generated. This paper proposes the parameters to select topic concepts, and describes the detailed algorithm of concept hierarchy tree building, concept counting and its application in summarizing. The experiment result shows that compared to word counting, this new method has preferably improved the performance of the system, and it helps to solve the abstract distribution problem of multi-topic texts.