Automatically Select Emotion for Response via Personality-affected Emotion Transition

Findings of the Association for Computational Linguistics: ACL-IJCNLP 2021, pages 5010–5020 August 1–6, 2021. ©2021 Association for Computational Linguistics 5010 Automatically Select Emotion for Response via Personality-affected Emotion Transition Zhiyuan Wen, Jiannong Cao, Ruosong Yang, Shuaiqi Liu, Jiaxing Shen Department of Computing, The Hong Kong Polytechnic University Kowloon, Hong Kong, China {cszwen, csjcao, csryang, cssliu, jiaxshen}@comp.polyu.edu.hk Abstract To provide consistent emotional interaction with users, dialog systems should be capable to automatically select appropriate emotions for responses like humans. However, most ex- isting works focus on rendering specified emo- tions in responses or empathetically respond to the emotion of users, yet the individual dif- ference in emotion expression is overlooked. This may lead to inconsistent emotional ex- pressions and disinterest users. To tackle this issue, we propose to equip the dialog system with personality and enable it to automatically select emotions in responses by simulating the emotion transition of humans in conversation. In detail, the emotion of the dialog system is transitioned from its preceding emotion in con- text. The transition is triggered by the preced- ing dialog context and affected by the spec- ified personality trait. To achieve this, we first model the emotion transition in the dia- log system as the variation between the pre- ceding emotion and the response emotion in the Valence-Arousal-Dominance (VAD) emo- tion space. Then, we design neural networks to encode the preceding dialog context and the specified personality traits to compose the vari- ation. Finally, the emotion for response is selected from the sum of the preceding emo- tion and the variation. We construct a dia- log dataset with emotion and personality la- bels and conduct emotion prediction tasks for evaluation. Experimental results validate the effectiveness of the personality-affected emo- tion transition.1. 1 Introduction Emotional intelligence can be considered a mental ability to reason validly with emotional informa- tion, and the action of emotions to enhance thought (Mayer, 2004). Hence, to create dialog systems with emotional intelligence during communication, 1Our dataset is released at: github.com/preke/PELD it is necessary to enable the machine to understand the emotion of users, select appropriate response emotions and express in conversation. Existing works either focus on rendering spec- ified emotions in responses (Zhou et al., 2018; Colombo et al., 2019), or understanding the emo- tion of users and respond empathetically (Zandie and Mahoor, 2020; Zhong et al., 2020; Lin et al., 2019); but how to automatically select the emo- tion for response is seldom discussed. Wei et al. (2019) proposes to learn appropriate emotional re- sponses from massive anonymous online dialogues. However, trained on conversations from different speakers, the dialog system ignores the individual difference of expressing emotions. This may lead to inconsistent emotional interactions and disinter- est users as they may feel they are still talking to rigid machines. In a dialog system, automatically selecting the emotion for response is to decide an emotion to be expressed facilitating the emotional response generation. Emotion selection can be modeled as the emotion transition (Thornton and Tamir, 2017), which refers to …

Content truncated to 500 words