基于多尺度自注意力增强的多方对话角色识别方法

Multi-party Dialogue Character Identification Method Based on Multi-scale Self-attention Enhancement

角色识别任务是近年来提出的一项自然语言处理任务,面向多方参与的对话场景,目标是将对话中的人物提及映射到具体的人物实体。目前在该任务的最优系统中,只使用了较为简单的编码器,并未针对对话文本特点进行改造创新。该文在最优系统的基础上,提出了一种基于多尺度自注意力增强的方法,借助不同尺度的自注意力,来获得更好的信息表示。首先,通过尺度较大的全局注意力,对场景内的全部对话信息进行处理,保留了全局的对话信息;然后,通过尺度较小的局部注意力,对局部范围内的对话进行计算,捕获近距离的信息之间的关联关系;最后,将不同尺度得到的信息进行融合,达到对编码信息增强的效果。在SemEval2018 Task4任务上的实验结果表明了该方法的有效性,相较于目前最优系统,在全部实体的F1值上提高了18.94%。

The character identification task aims at mapping the person mentions in the dialogue to specific person entities in the dialogue scenarios involving multiple parties. This paper proposes a method based on multi-scale self-attention enhancement, which uses self-attention at different scales to obtain better information representation. First, the global dialog information in the scene is captured through global attention with a large scope. Then, through the small-scale local attention, the dialog in the local area is calculated to capture the association relationship between the information at close range. Finally, the information obtained at different scales is fused to enhance the encoded information. The experimental results on SemEval2018 Task4 show the effectiveness of the method by 18.94% in F1 compared with the current optimal system.