A Discussion on Various Methods in Automatic Abstractive Text Summarization

Automatic abstractive text summarization (ATS) is important in terms of optimized data storage and transmission requirements on technical aspect and also for quick view of the information for the people with their busy lives. This perspective and applicability have triggered the thoughts of various researchers to develop the efficient models that can provide consistently good in terms of understandable, meaningful and short summaries. The main aim of this discussion is to address the literature available in the domain for understanding the researcher’s perspective of development of such models and to form the platform of thoughts for the researchers in the field to finalize new strategies of the development. Some of the papers addressed here show the use of neural network platform in ATS along with graph-based hypothetical methods which show significant difference in perspectives considered while developing the models and flexibility and complexity involved in the implementation for the sake of experimentation and evaluating the performance of models.