Observing Dialogue in Therapy: Categorizing and Forecasting Behavioral Codes

Proceedings of the 57th Annual Meeting of the Association for Computational Linguistics, pages 5599–5611 Florence, Italy, July 28 - August 2, 2019. c⃝2019 Association for Computational Linguistics 5599 Observing Dialogue in Therapy: Categorizing and Forecasting Behavioral Codes Jie Cao†, Michael Tanana‡, Zac E. Imel‡, Eric Poitras‡, David C. Atkins♦, Vivek Srikumar† †School of Computing, University of Utah ‡Department of Educational Psychology, University of Utah ♦Department of Psychiatry and Public Health, University of Washington {jcao, svivek}@cs.utah.edu, {michael.tanana, zac.imel, eric.poitras}@utah.edu, datkins@u.washington.edu Abstract Automatically analyzing dialogue can help un- derstand and guide behavior in domains such as counseling, where interactions are largely mediated by conversation. In this paper, we study modeling behavioral codes used to asses a psychotherapy treatment style called Motiva- tional Interviewing (MI), which is effective for addressing substance abuse and related prob- lems. Specifically, we address the problem of providing real-time guidance to therapists with a dialogue observer that (1) categorizes therapist and client MI behavioral codes and, (2) forecasts codes for upcoming utterances to help guide the conversation and potentially alert the therapist. For both tasks, we define neural network models that build upon recent successes in dialogue modeling. Our experi- ments demonstrate that our models can outper- form several baselines for both tasks. We also report the results of a careful analysis that re- veals the impact of the various network design tradeoffs for modeling therapy dialogue. 1 Introduction Conversational agents have long been studied in the context of psychotherapy, going back to chat- bots such as ELIZA (Weizenbaum, 1966) and PARRY (Colby, 1975). Research in modeling such dialogue has largely sought to simulate a par- ticipant in the conversation. In this paper, we argue for modeling dialogue observers instead of participants, and focus on psychotherapy. An observer could help an ongo- ing therapy session in several ways. First, by mon- itoring fidelity to therapy standards, a helper could guide both veteran and novice therapists towards better patient outcomes. Second, rather than gen- erating therapist utterances, it could suggest the type of response that is appropriate. Third, it could alert a therapist about potentially important cues from a patient. Such assistance would be espe- cially helpful in the increasingly prevalent online or text-based counseling services.1 We ground our study in a style of therapy called Motivational Interviewing (MI, Miller and Roll- nick, 2003, 2012), which is widely used for treat- ing addiction-related problems. To help train ther- apists, and also to monitor therapy quality, ut- terances in sessions are annotated using a set of behavioral codes called Motivational Interviewing Skill Codes (MISC, Miller et al., 2003). Table 1 shows standard therapist and patient (i.e., client) codes with examples. Recent NLP work (Tanana et al., 2016; Xiao et al., 2016; P´erez-Rosas et al., 2017; Huang et al., 2018, inter alia) has studied the problem of using MISC to assess completed sessions. Despite its usefulness, automated post hoc MISC labeling does not address the desiderata for ongoing sessions identified above; such mod- els use information from utterances yet to be …

Content truncated to 500 words