Text Extraction and Finetuning Transformers for Abstractive Summarisation

Data has always been an integral part of any industry that people refer to nowadays. And due to the advancement of technology more and more people have been able to access data and generate it effortlessly. This often leads to generation of large text documents which often times take a lot of energy. Natural Language Processing has been a constantly evolving field as it helps to summarize the long texts and allow people to get the main data in a short and concise manner thereby saving time and energy. In this paper an Abstractive Text Summarizer built on the concepts of NLP has been proposed. The proposed method has been implemented on two major and well known domains - News Articles and Legal documents. The system proposed in this paper consists of three major models developed and finetuned on the datasets gathered from the domains mentioned above.