The goal of multi-document summarization is to give a prejudiced summary on a particular topic. This paper describes how to perform summarization task by means of multiple-ranking of sentences from various documents. Our method utilizes the association among all the sentences in the documents and the association between the sentences and the particular query. For each sentence in the documents, ranking score has been computed to denote the richness of information in the sentences. We used greedy algorithm to force multiple range of consequences on each sentence. By selecting the sentences with highly prejudiced information richness and high information originality, the summary has been produced. We make use of both TREC and TDT documents for our experimental purpose and our results show that the proposed methodology outperforms well when compared with the existing methodologies.
展开▼
