A Computational Approach to Understanding Empathy Expressed in Text-Based Mental Health Support

Empathy is critical to successful mental health support. Empathy measurement has predominantly occurred in synchronous, face-to-face settings, and may not translate to asynchronous, text-based contexts. Because millions of people use text-based platforms for mental health support, understanding empathy in these contexts is crucial. In this work, we present a computational approach to understanding how empathy is expressed in online mental health platforms. We develop a novel unifying theoretically-grounded framework for characterizing the communication of empathy in text-based conversations. We collect and share a corpus of 10k (post, response) pairs annotated using this empathy framework with supporting evidence for annotations (rationales). We develop a multi-task RoBERTa-based bi-encoder model for identifying empathy in conversations and extracting rationales underlying its predictions. Experiments demonstrate that our approach can effectively identify empathic conversations. We further apply this model to analyze 235k mental health interactions and show that users do not self-learn empathy over time, revealing opportunities for empathy training and feedback.

[1]  Munmun De Choudhury,et al.  Mental Health Support and its Relationship to Linguistic Accommodation in Online Communities , 2018, CHI.

[2]  Michael B. Roberts,et al.  Curricula for empathy and compassion training in medical education: A systematic review , 2019, PloS one.

[3]  Mark H. Davis,et al.  A Multidimensional Approach to Individual Differences in Empathy , 1980 .

[4]  Jakob Grue Simonsen,et al.  A Hierarchical Recurrent Encoder-Decoder for Generative Context-Aware Query Suggestion , 2015, CIKM.

[5]  Rosalind W. Picard,et al.  Crowdsourcing Collective Emotional Intelligence , 2012, ArXiv.

[6]  G. Barrett‐Lennard,et al.  The empathy cycle: Refinement of a nuclear concept. , 1981 .

[7]  Anthony P Morrison,et al.  The Lancet Psychiatry Commission on psychological treatments research in tomorrow's science. , 2018, Lancet psychiatry.

[8]  Jason Weston,et al.  Poly-encoders: Transformer Architectures and Pre-training Strategies for Fast and Accurate Multi-sentence Scoring , 2019 .

[9]  Lukasz Kaiser,et al.  Attention is All you Need , 2017, NIPS.

[10]  Zheng Yao,et al.  The Channel Matters: Self-disclosure, Reciprocity and Social Support in Online Cancer Support Groups , 2019, CHI.

[11]  Verónica Pérez-Rosas,et al.  What Makes a Good Counselor? Learning to Distinguish between High-quality and Low-quality Counseling Conversations , 2019, ACL.

[12]  Amit Sharma,et al.  Moments of Change: Analyzing Peer-Based Cognitive Support in Online Mental Health Forums , 2019, CHI.

[13]  Verónica Pérez-Rosas,et al.  Understanding and Predicting Empathic Behavior in Counseling Therapy , 2017, ACL.

[14]  J. Watson,et al.  Client-centered and experiential psychotherapy in the 21st century : advances in theory, research and practice , 2002 .

[15]  Kathleen McKeown,et al.  Identifying therapist conversational actions across diverse psychotherapeutic approaches , 2019, Proceedings of the Sixth Workshop on Computational Linguistics and Clinical Psychology.

[16]  Alan E. Kazdin,et al.  Novel Models for Delivering Mental Health Services and Reducing the Burdens of Mental Illness , 2013 .

[17]  Jason Whipple,et al.  Do psychotherapists improve with time and experience? A longitudinal analysis of outcomes in a clinical setting. , 2016, Journal of counseling psychology.

[18]  M. White,et al.  Receiving social support online: implications for health education. , 2001, Health education research.

[19]  Adam S. Miner,et al.  Key Considerations for Incorporating Conversational AI in Psychotherapy , 2019, Front. Psychiatry.

[20]  Cristian Danescu-Niculescu-Mizil,et al.  Finding Your Voice: The Linguistic Development of Mental Health Counselors , 2019, ACL.

[21]  Omer Levy,et al.  RoBERTa: A Robustly Optimized BERT Pretraining Approach , 2019, ArXiv.

[22]  Y-Lan Boureau,et al.  Towards Empathetic Open-domain Conversation Models: A New Benchmark and Dataset , 2018, ACL.

[23]  Cornelia Caragea,et al.  Identifying Empathetic Messages in Online Health Communities , 2017, IJCNLP.

[24]  Ming-Wei Chang,et al.  BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding , 2019, NAACL.

[25]  R. Selman.  growth of interpersonal understanding , 1980 .

[26]  Doug Downey,et al.  Don’t Stop Pretraining: Adapt Language Models to Domains and Tasks , 2020, ACL.

[27]  Mark Olfson,et al.  Building The Mental Health Workforce Capacity Needed To Treat Adults With Serious Mental Illnesses. , 2016, Health affairs.

[28]  John C. Norcross,et al.  Psychotherapy relationships that work : therapist contributions and responsiveness to patients , 2002 .

[29]  Robert R. Carkhuff,et al.  Toward effective counseling and psychotherapy : training and practice , 2017 .

[30]  R. Hayward.  Empathy , 1971, The Lancet.

[31]  Lyle H. Ungar,et al.  Modeling Empathy and Distress in Reaction to News Stories , 2018, EMNLP.

[32]  Modern Applications of Psychology , 1937, Nature.

[33]  Ilya Sutskever,et al.  Language Models are Unsupervised Multitask Learners , 2019 .

[34]  Zijian Wang,et al.  It’s going to be okay: Measuring Access to Support in Online Communities , 2018, EMNLP.

[35]  Yejin Choi,et al.  Social Bias Frames: Reasoning about Social and Power Implications of Language , 2020, ACL.

[36]  J. Powell,et al.  Primary care Health related virtual communities and electronic support groups : systematic review of the effects of online peer to peer interactions , 2004 .

[37]  Panayiotis G. Georgiou,et al.  A Deep Learning Approach to Modeling Empathy in Addiction Counseling , 2016, INTERSPEECH.

[38]  Jianfeng Gao,et al.  Challenges in Building Intelligent Open-domain Dialog Systems , 2019, ACM Trans. Inf. Syst..

[39]  Byron C. Wallace,et al.  ERASER: A Benchmark to Evaluate Rationalized NLP Models , 2020, ACL.

[40]  David C. Atkins,et al.  More than reflections: empathy in motivational interviewing includes language style synchrony between therapist and client. , 2015, Behavior therapy.

[41]  Cristian Danescu-Niculescu-Mizil,et al.  Balancing Objectives in Counseling Conversations: Advancing Forwards or Looking Backwards , 2020, ACL.

[42]  Jianfeng Gao,et al.  DialoGPT: Large-Scale Generative Pre-training for Conversational Response Generation , 2020, ACL.

[43]  Jure Leskovec,et al.  Large-scale Analysis of Counseling Conversations: An Application of Natural Language Processing to Mental Health , 2016, TACL.

[44]  Jian Sun,et al.  Deep Residual Learning for Image Recognition , 2015, 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR).

[45]  Robert Elliott,et al.  Therapist Empathy and Client Outcome: An Updated Meta-Analysis , 2018, Psychotherapy.

[46]  Tara Matthews,et al.  Stories from Survivors: Privacy & Security Practices when Coping with Intimate Partner Abuse , 2017, CHI.

[47]  David C. Atkins,et al.  Computational psychotherapy research: scaling up the evaluation of patient-provider interactions. , 2015, Psychotherapy.