Open Access

ARTICLE

A Method of Integrating Length Constraints into Encoder-Decoder Transformer for Abstractive Text Summarization

Ngoc-Khuong Nguyen^1,2, Dac-Nhuong Le¹, Viet-Ha Nguyen², Anh-Cuong Le^3,*

1 Faculty of Information Technology, Haiphong University, Haiphong, 180000, Vietnam
2 Faculty of Information Technology, VNU University of Engineering and Technology, Hanoi, 10000, Vietnam
3 Natural Language Processing and Knowledge Discovery Laboratory, Faculty of Information Technology, Ton Duc Thang University, Ho Chi Minh City, Vietnam

* Corresponding Author: Anh-Cuong Le. Email:

Intelligent Automation & Soft Computing 2023, 38(1), 1-18. https://doi.org/10.32604/iasc.2023.037083

Received 22 October 2022; Accepted 20 February 2023; Issue published 26 January 2024

Abstract

Text summarization aims to generate a concise version of the original text. The longer the summary text is, the more detailed it will be from the original text, and this depends on the intended use. Therefore, the problem of generating summary texts with desired lengths is a vital task to put the research into practice. To solve this problem, in this paper, we propose a new method to integrate the desired length of the summarized text into the encoder-decoder model for the abstractive text summarization problem. This length parameter is integrated into the encoding phase at each self-attention step and the decoding process by preserving the remaining length for calculating head-attention in the generation process and using it as length embeddings added to the word embeddings. We conducted experiments for the proposed model on the two data sets, Cable News Network (CNN) Daily and NEWSROOM, with different desired output lengths. The obtained results show the proposed model’s effectiveness compared with related studies.

---

Keywords

---