Document summarization method based on heterogeneous graph

Document summarization has been widely studied for many years. Existing methods mainly use statistical or linguistic information to extract the most informative sentences from document. However, those methods ignore the relationship between different granularities (i.e., word, sentence, and topic). Actually, the interactions between those granularities can be used in document summarization. In this paper we proposed a document summarization method based on heterogeneous graph. The method is first implemented by constructing a graph which reflect relationship between different size of granularity nodes, and then using ranking algorithm to calculate score of nodes. Finally, highest score of sentences in the document will be chosen as summary. Experimental results show that our approach outperforms baseline methods.