VALUE: Understanding Dialect Disparity in NLU

Proceedings of the 60th Annual Meeting of the Association for Computational Linguistics Volume 1: Long Papers, pages 308 - 319 May 22-27, 2022 c⃝2022 Association for Computational Linguistics MISC: A MIxed Strategy-Aware Model Integrating COMET for Emotional Support Conversation Quan Tu1∗ † , Yanran Li2∗, Jianwei Cui2, Bin Wang2, Ji-Rong Wen1,3 and Rui Yan1,3‡ 1Gaoling School of Artificial Intelligence, Renmin University of China 2Xiaomi AI Lab 3Beijing Academy of Artificial Intelligence 1{quantu,jrwen,ruiyan}@ruc.edu.cn 2{liyanran,cuijianwei,wangbin11}@xiaomi.com Abstract Applying existing methods to emotional sup- port conversation—which provides valuable as- sistance to people who are in need—has two major limitations: (a) they generally employ a conversation-level emotion label, which is too coarse-grained to capture user’s instant men- tal state; (b) most of them focus on expressing empathy in the response(s) rather than gradu- ally reducing user’s distress. To address the problems, we propose a novel model MISC, which firstly infers the user’s fine-grained emo- tional status, and then responds skillfully us- ing a mixture of strategy. Experimental re- sults on the benchmark dataset demonstrate the effectiveness of our method and reveal the benefits of fine-grained emotion understand- ing as well as mixed-up strategy modeling. Our code and data could be found in https: //github.com/morecry/MISC. 1 Introduction Empathy is the ability to perceive what others feel, think in their places and respond properly. It has a broad application scenarios to endow machines with the ability of empathy, including automatic psycho-therapist, intelligent customer service, em- pathetic conversational agents, and etc (Fitzpatrick et al., 2017; Shin et al., 2019; Ma et al., 2020). In this work, we focus on a special kind of human-computer empathetic conversation, i.e., emotional support conversation (Liu et al., 2021). Distinguishedly, emotional support conversation happens between a seeker and supporter, where the supporter aims to gradually reduce seeker’s distress as the conversation goes. This makes ex- isting approaches unsuitable for our setting for at least two reasons. Firstly, existing work on emo- tional chatting learns to predict user emotion us- ing a conversation-level emotion label, which is ∗Equal Contribution. †This work was done during internship at Xiaomi AI Lab. ‡Corresponding author: Rui Yan (ruiyan@ruc.edu.cn). coarse-grained and static to the conversation con- text (Rashkin et al., 2019; Lin et al., 2019c; Li et al., 2020a). However, emotion is complex and user emotion intensity will change during the de- veloping of the conversation (Liu et al., 2021). It is thus a necessity to tell seeker’s fine-grained mental state at each utterance. Secondly, most of empa- thetic chatbots are trained to respond emotionally in accordance with the predicted coarse-grained emotion class, without consideration on how to ad- dress the seeker’s emotional problem (De Graaf et al., 2012; Majumder et al., 2020; Xie and Park, 2021). Hence, they are deficient to apply for emo- tional support conversation whose goal is to help others work through the challenges they face. Figure 1: An Emotional Support Conversation Example. To tackle these issues, we propose a novel ap- proach MISC, a.k.a. MIxed Srategy-aware model integrating COMET for emotional support conver- sation. …

Content truncated to 500 words