arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2203.07216cs.CL
收藏

看待注意力的新视角:面向新闻分类的基于双层注意力的可解释主题建模

A Novel Perspective to Look At Attention: Bi-level Attention-based Explainable Topic Modeling for News Classification

Insight Centre for Data Analytics(洞察数据分析中心);University College Dublin(都柏林大学学院)

Dairui Liu, Derek Greene, Ruihai Dong

AI总结:

针对深度学习注意力机制在NLP任务中可解释性不足的问题,提出基于双层注意力架构的可解释主题建模框架,将解释与决策解耦,在新闻分类任务上性能达最优水平且具备良好可解释性。

摘要:

近年来,许多基于深度学习的方案已在自然语言处理(NLP)领域的各类任务中广泛采用注意力机制。然而,深度学习模型的固有特性与注意力机制的灵活性提升了模型复杂度,进而给模型可解释性带来挑战。本文针对这一挑战,提出一种新颖实用的框架,采用双层注意力架构,将解释复杂度与决策过程解耦。我们将其应用于新闻文章分类任务场景。在两个大规模新闻语料库上的实验表明,所提模型可与多种现有最优(state-of-the-art)方案取得相当的性能,并从可解释性角度验证了其适用性。

英文摘要:

Many recent deep learning-based solutions have widely adopted the attention-based mechanism in various tasks of the NLP discipline. However, the inherent characteristics of deep learning models and the flexibility of the attention mechanism increase the models' complexity, thus leading to challenges in model explainability. In this paper, to address this challenge, we propose a novel practical framework by utilizing a two-tier attention architecture to decouple the complexity of explanation and the decision-making process. We apply it in the context of a news article classification task. The experiments on two large-scaled news corpora demonstrate that the proposed model can achieve competitive performance with many state-of-the-art alternatives and illustrate its appropriateness from an explainability perspective.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑