Graph based Extractive Multi-document Summarizer for Malayalam-an Experiment

Multidocument summarization is an automatic process to generate summary extract from multiple documents written about the same topic. Of the many summarization systems developed for English language, the graph based system is found to be more effective. This paper mainly focuses on a multidocument summarizing system for Malayalam Language which follows a graph based approach. The proposed model uses a weighted undirected graph to represent the documents. The significant sentences for the summary are selected by applying the Page Rank algorithm. Experimental results demonstrate the effectiveness of the proposed system.