Int J Performability Eng ›› 2021, Vol. 17 ›› Issue (10): 900-906.doi: 10.23940/ijpe.21.10.p8.900906

Previous Articles    

Fine-Tuned T5 for Abstractive Summarization

Abdul Ghafoor Etemad*, Ali Imam Abidi, and Megha Chhabra   

  1. Department of Computer Science & Engineering, School of Engineering & Technology, Sharda University, Greater Noida, 201310, India
  • Submitted on ; Revised on ; Accepted on
  • Contact: *E-mail address: ghafooretemad3@gmail.com
  • About author:Abdul Ghafoor Etemad is an aspiring scholar who is pursuing a master’s degree from Department of Computer Science & Engineering, School of Engineering & Technology, Sharda University, Greater Noida, India. His research interests include Data Summarization etc.
    Ali Imam Abidi is an Assistant Professor at the Department of Computer Science & Engineering, School of Engineering & Technology, Sharda University, Greater Noida, India. His research interests include Computer Vision, Feature Data Analysis etc.
    Megha Chhabra is Assistant Professor at the Department of Computer Science & Engineering, School of Engineering & Technology, Sharda University, Greater Noida, India. Her research interests include Machine Learning, Image Forensics etc.

Abstract: Abstract Text Summarization can be understood as the task of constructing a summary from a relatively larger text. This summary would comprise of only a comparatively much smaller number of sentences than the actual text and would still express the main idea. Its applications lie in sentiment analysis, document summarization, search engine queries, business analysis, etc. Over time, a lot of research has happened on the topic of abstract text summarization, especially with the emergence of pre-trained models proposed by researchers. In this research a pre-trained model was fine-tuned on Xsum and Gigaword datasets and produced state-of-the-art performance in the abstractive summarization.

Key words: abstract text summarization, extract text summarization, convolutional neural network, recurrent neural network, sequence-to-sequence model, transformer, encoder-decoder, attention mechanism