A Mixed Approach for Gist Generation of Hindi News Article

Text summarization systems generates summary of articles that makes users to quickly identify the theme of the retrieved articles. Summary can be viewed as gists, table of contents, outlines, minutes, previews, abstracts, synopses, reviews, digests, abridgments, biographies, bulletins, histories and sound bites. gist of an article offers important clues that can help the reader to quickly decide whether to skip, to scan or to read the article. An attempt is made to extract a generic text summary that identifies topic sentences of Hindi text article. Our method aims at to rank the sentences in the article by using different surface features and extract the higher ranked sentences in order to generate a summary with an extensive coverage of the main content of the news article. Various evaluation techniques are used to compare the machine generated output with human generated summaries.