Comprehensive Survey on Abstractive Text Summarization

Abstract

Natural language generation process is currently utilized in the field of Machine translation, scientific article summarization, web-blogs and other social media platforms. Natural language generation is nearly an abstractive text summarization approach which ingests the complex information content in various contexts and presents user's apportioned summary of length in the precise manner. Concise informative summaries generated should conserve valid information free from noise. Lossless significant information extraction, and concrete relation assessment and concept labelling is the most pivotal task in abstractive summary production. This study investigates existing techniques applied for abstractive text summarization and its significance. In addition, we explore the possibility of ontology based multi document abstract summarization with valuable adoption of conventional approaches to uplift the abstractive text summarization.