An improved web information summarization based on SSSC

This paper proposed a new method of web news summarization via soft clustering algorithm. It used search engine to extract relevant documents, and mixed query sentence into sentences set which segmented from multi-document set, then this paper adopted efficient soft cluster algorithm SSSC to cluster all the sentences. If the number of cluster which contains the query sentence is larger than or equal to 5, the summary sentence will be extracted by turns from the clusters which query sentence in, or feature fusion will be used to extract summary sentences. Experimental result shows that the proposed summarization method can improve the performance of summary, soft clustering algorithm is efficient.