A Novel Features Based Automated Gurmukhi Text Summarization System

Abstract

Today it is very difficult, laborious and time consuming task to extract information manually from large amount of data available. A need of Automatic Text Summarization Tool is growing which will create complete and understandable summarized text of input text document. Many summarization techniques have been developed for English language but very less research work is carried out in the area of Punjabi Text Summarization. Mostly Gurmukhi script is used for writing Punjabi language. In this paper, the idea to summarize Gurmukhi text document using extraction technique is discussed which summarized text based on features extracted from document. New features namely Presence of URL’s or Email Addresses, Presence of Brackets and Presence of Inverted Commas in sentences have been analyzed, proposed and compared with some older features available for text summarization in literature. The document summary generated by proposed system has been compared with summary of document generated by human experts. The comparison shows better results in the terms of Precision, Recall and F-score.