BiSparse-AAS: Bilinear Sparse Attention and Adaptive Spans Framework for Scalable and Efficient Text Summarization

Fuente: arXiv
Enregistré dans:
Détails bibliographiques
Auteurs principaux: Hagos, Desta Haileselassie, Burge, Legand L., Andy, Anietie, Yazidi, Anis, Vlassov, Vladimir
Format: Preprint
Publié: 2025
Sujets:
Accès en ligne:
Tags: Ajouter un tag
Pas de tags, Soyez le premier à ajouter un tag!
_version_ 1866915588625924096
author Hagos, Desta Haileselassie
Burge, Legand L.
Andy, Anietie
Yazidi, Anis
Vlassov, Vladimir
author_facet Hagos, Desta Haileselassie
Burge, Legand L.
Andy, Anietie
Yazidi, Anis
Vlassov, Vladimir
contents Transformer-based architectures have advanced text summarization, yet their quadratic complexity limits scalability on long documents. This paper introduces BiSparse-AAS (Bilinear Sparse Attention with Adaptive Spans), a novel framework that combines sparse attention, adaptive spans, and bilinear attention to address these limitations. Sparse attention reduces computational costs by focusing on the most relevant parts of the input, while adaptive spans dynamically adjust the attention ranges. Bilinear attention complements both by modeling complex token interactions within this refined context. BiSparse-AAS consistently outperforms state-of-the-art baselines in both extractive and abstractive summarization tasks, achieving average ROUGE improvements of about 68.1% on CNN/DailyMail and 52.6% on XSum, while maintaining strong performance on OpenWebText and Gigaword datasets. By addressing efficiency, scalability, and long-sequence modeling, BiSparse-AAS provides a unified, practical solution for real-world text summarization applications.
format Preprint
id arxiv_https___arxiv_org_abs_2510_27516
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle BiSparse-AAS: Bilinear Sparse Attention and Adaptive Spans Framework for Scalable and Efficient Text Summarization
Hagos, Desta Haileselassie
Burge, Legand L.
Andy, Anietie
Yazidi, Anis
Vlassov, Vladimir
Computation and Language
Machine Learning
Transformer-based architectures have advanced text summarization, yet their quadratic complexity limits scalability on long documents. This paper introduces BiSparse-AAS (Bilinear Sparse Attention with Adaptive Spans), a novel framework that combines sparse attention, adaptive spans, and bilinear attention to address these limitations. Sparse attention reduces computational costs by focusing on the most relevant parts of the input, while adaptive spans dynamically adjust the attention ranges. Bilinear attention complements both by modeling complex token interactions within this refined context. BiSparse-AAS consistently outperforms state-of-the-art baselines in both extractive and abstractive summarization tasks, achieving average ROUGE improvements of about 68.1% on CNN/DailyMail and 52.6% on XSum, while maintaining strong performance on OpenWebText and Gigaword datasets. By addressing efficiency, scalability, and long-sequence modeling, BiSparse-AAS provides a unified, practical solution for real-world text summarization applications.
title BiSparse-AAS: Bilinear Sparse Attention and Adaptive Spans Framework for Scalable and Efficient Text Summarization
topic Computation and Language
Machine Learning
url https://arxiv.org/abs/2510.27516