Research on Text Summarization Applications Based on Deep Learning

DOI

10.2991/978-94-6463-823-3_99How to use a DOI?

Keywords

Text Summarization; Deep Learning; Pre-trained Language Models; Large Language Models

Abstract

In the information age, the volume of text data has surged, making manual processing inefficient. Thus, text summarization technology is crucial. This paper reviews various techniques, from traditional statistic methods to large language models, and identifies their limitations. Traditional statistic methods are fast but fail to grasp semantics, producing incoherent summaries. Sequence to Sequence (Seq2Seq) models struggle with long texts and deviate from main topics. Transformer models rely heavily on lexical features, limiting cross-domain adaptability. Pre-trained language models lack flexibility in understanding complex semantics and generation strategies. Large language models often produce factual errors and have flawed dynamic importance assessment. To address these issues, this paper proposes improvements. Integrating traditional statistic methods with deep learning and semantic analysis; constructing a dynamic topic constraint mechanism for Seq2Seq models and introducing graph neural networks; designing a new attention module for Transformers and building a differentiable recombination architecture; assisting pre-trained language model training with knowledge graphs and adjusting model architecture and methods; optimizing large language model generation by combining reinforcement learning with knowledge reasoning. These enhancements aim to improve text summarization quality, ensuring accurate core information extraction, coherence, logical consistency, adaptability, reduced errors, and better application across multiple fields.

Copyright

© 2025 The Author(s)

Open Access

Open Access This chapter is licensed under the terms of the Creative Commons Attribution-NonCommercial 4.0 International License (http://creativecommons.org/licenses/by-nc/4.0/), which permits any noncommercial use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license and indicate if changes were made.

Download article (PDF)

TY  - CONF
AU  - Zixuan Yang
PY  - 2025
DA  - 2025/08/31
TI  - Research on Text Summarization Applications Based on Deep Learning
BT  - Proceedings of the 2025 3rd International Conference on Image, Algorithms, and Artificial Intelligence (ICIAAI 2025)
PB  - Atlantis Press
SP  - 1016
EP  - 1028
SN  - 2352-538X
UR  - https://doi.org/10.2991/978-94-6463-823-3_99
DO  - 10.2991/978-94-6463-823-3_99
ID  - Yang2025
ER  -