Automatic Text Summarization from Unstructured Text using Natural Language Processing

Abstract

Text summarization is one which creates short, concrete, fluent data from a longer document. Text summarization is of two types: one is abstractive, and another is extractive. In this paper, automatic text summarization is done using extractive summarization. For the idea of generating deserved summary, the unstructured text is first pre-processing which includes -tokenization, parts of speech, chunking. After pre-processing, feature extraction process take place which uses scoring method and a novel reinforcement learning based training algorithm to extract the data, which is greater than threshold value, that extract data is stored for final summary.