Article: Automatic summarisation of digital media news based on the transformer model Journal: International Journal of Information and Communication Technology (IJICT) 2025 Vol.26 No.25 pp.1 - 18 Abstract: The increasing expansion of digital media content presents a significant difficulty in the field of information processing since it becomes difficult to effectively create accurate and succinct summaries from large news sources. Especially with complex and multimodal news content, traditional news summary generating techniques are sometimes difficult to consider the information coverage; the impact is restricted. This work so suggests, based on the transformer model, an automatic summarising method for digital media news. While decreasing the development of repetitive content via redundancy penalty factors, sentence vector augmentation and keyword guiding techniques help to more precisely capture the relevant information in the news. The strategy suggested in this work greatly beats the conventional summary generating model in the ROUGE series of metrics. New technical solutions and value references for automated processing and intelligent generation of digital media news are presented by this work. Inderscience Publishers - linking academia, business and industry through research

Title: Automatic summarisation of digital media news based on the transformer model

Authors: Xuan Wu; Guoyan Zhang

Addresses: Xi'an Jiaotong University City College, Xi'an 710018, China ' Krirk University, Bangkok 10220, Thailand

Abstract: The increasing expansion of digital media content presents a significant difficulty in the field of information processing since it becomes difficult to effectively create accurate and succinct summaries from large news sources. Especially with complex and multimodal news content, traditional news summary generating techniques are sometimes difficult to consider the information coverage; the impact is restricted. This work so suggests, based on the transformer model, an automatic summarising method for digital media news. While decreasing the development of repetitive content via redundancy penalty factors, sentence vector augmentation and keyword guiding techniques help to more precisely capture the relevant information in the news. The strategy suggested in this work greatly beats the conventional summary generating model in the ROUGE series of metrics. New technical solutions and value references for automated processing and intelligent generation of digital media news are presented by this work.

Keywords: digital media news; automatic summarisation; transformer model; multimodal information; redundancy penalty; optimisation strategy.

DOI: 10.1504/IJICT.2025.147138

International Journal of Information and Communication Technology, 2025 Vol.26 No.25, pp.1 - 18

Received: 01 May 2025
Accepted: 10 May 2025
Published online: 10 Jul 2025 *

Title: Automatic summarisation of digital media news based on the transformer model

Keep up-to-date