Extractive text summarization by feature-based sentence extraction using rule-based concept

World Wide Web is a tremendous source of knowledge. Vast amount of information available over the internet has made the humans suffer from problem of information explosion. Therefore, a good mechanism is required to extract relevant information. This calls a need for an automatic and significant tool that converts lengthy documents into concise form by extracting relevant information from it. Automatic Text Summarizer serves as one of the best tool for interpreting lengthy textual content. It represents the shorter version of the original document by choosing most important part of text, thus generating its summary. It is classified into two categories: abstraction and extraction. This paper highlights on extractive approach. Main aim is to select best sentences by weighting them. We carried out our experiment on 15 documents from DUC 2002 dataset. Each test document was first pre-processed. Then, all the sentences were represented as attribute vector of features by calculating their scores. Rule-based method was proposed to select the best sentences. Results were compared with GSM summarizer and conclusion was drawn that best average recall, precision and f-measure values was obtained for Rule-Based Summarizer.