实验室关于情绪原因提取任务研究被期刊Neural Computing and Application录用
新闻来源:IR实验室       发布时间:2019/6/28 15:25:24

  近日,收到期刊《Neural Computing and Application》编辑部邮件,实验室刁宇峰等的研究工作“Multi-Granularity Bi-Directional Attention Stream Machine Comprehension Method For Emotion Cause Extraction”已被录用。

  摘要:情绪原因提取任务是指抽取能够触发人们情绪表达的原因信息。该任务在自然语言处理应用中处于关键的研究地位,如情感分析和语义理解系统等。但是,现有的大多数情绪原因提取算法大多集中在特征工程方面,往往忽视了情绪词和其上下文之间的潜在语义信息。在本文,我们提出一种新颖的可计算的基于阅读理解架构的Multi-Granularity Bi-Directional Attention Stream(MBiAS)模型来解决该问题。Context和Query分别基于细粒度的词嵌入模型进行建模。双向注意力流机制被用于获取情绪Query方面的Context表示向量。此外,我们在开放中文情绪原因数据集上进行丰富的实验,其实验结果表明我们的方法已超越目前最先进的水平。

  Abstract:Emotion cause extraction is to extract the cause information that triggers the emotion expression of described person. This task in text plays a critical role in Nature Language Processing (NLP) applications, such as Sentiment Analysis (SA) and Semantic Comprehension system. However, most existing methods for this emotion cause extraction task only focus on feature engineering, and ignore the latent semantic information between emotion word and context to hinder the performance. In this paper, we propose a novel computational Multi-granularity Bi-Directional Attention Stream (MBiAS) network based on a machine comprehension frame to settle this problem. The context and query are embedded by this multi-stage hierarchical process based on the fine-grained levels of embeddings. Then the bi-directional attention stream mechanism is applied to get an emotional query-aware context representation. Meanwhile, we have conducted extensive experiments on available Chinese emotion cause dataset. The experimental results demonstrate that our approach significantly outperforms the state-of-the-art methods, and is able to extract the emotion cause.