Automatic Hindi Text Summarization Using Selection and Elimination Approach

In recent years, the shoot up of web information has required to make intensive research in the field of automatic text summarization which is the part of the Natural Language Processing (NLP). This paper concentrates on different algorithms for text summarization of Hindi documents and to check which methods among them, are given the best summary. This paper has used selection and elimination based approach in order to make a summarizer for Hindi text documents. To make final summary several features are taken into account, including sentence resemblance to the first 100 and last 100 words and find sentence frequency on the basis of the word frequency, and sentence similarity by using Dice Coefficient and cluster the summary obtained by both methods and the Jaccard's Coefficient of dissimilarity is used to remove the sentences which are highly dissimilar in the summary. The results of the proposed algorithm are compared over different algorithms used for text summarization using Hindi corpus posted by IIT Bombay to identify which approach contains the most important points of the original text.