From Words and Exercises to Wellness: Farsi Chatbot for Self-Attachment Technique

TL;DR

This study tackles the lack of mental health resources for Farsi speakers by building a voice-capable Self-Attachment Technique chatbot that blends rule-based dialogue with neural methods. It introduces a 12-class Farsi emotion classifier achieving $92.13\%$ accuracy on a 6,168-utterance dataset and uses PPO-trained ParsGPT-2 to generate a large pool of empathetic responses (≈$152{,}000$ utterances) for retrieval-based dialogue, complemented by a SAT Teacher QA module with $80.12\%$ accuracy. The system, evaluated in a non-clinical feasibility study with $N=51$ over ten days, achieved generally positive engagement and well-being indicators, with SAT Teacher satisfaction near 7.4/10; qualitative feedback identified benefits and areas for improvement. By releasing three public Farsi datasets and demonstrating effective user interaction, the work advances accessibility and reproducibility for digital psychotherapy in low-resource languages and lays groundwork for future clinical research.

Abstract

In the wake of the post-pandemic era, marked by social isolation and surging rates of depression and anxiety, conversational agents based on digital psychotherapy can play an influential role compared to traditional therapy sessions. In this work, we develop a voice-capable chatbot in Farsi to guide users through Self-Attachment (SAT), a novel, self-administered, holistic psychological technique based on attachment theory. Our chatbot uses a dynamic array of rule-based and classification-based modules to comprehend user input throughout the conversation and navigates a dialogue flowchart accordingly, recommending appropriate SAT exercises that depend on the user's emotional and mental state. In particular, we collect a dataset of over 6,000 utterances and develop a novel sentiment-analysis module that classifies user sentiment into 12 classes, with accuracy above 92%. To keep the conversation novel and engaging, the chatbot's responses are retrieved from a large dataset of utterances created with the aid of Farsi GPT-2 and a reinforcement learning approach, thus requiring minimal human annotation. Our chatbot also offers a question-answering module, called SAT Teacher, to answer users' questions about the principles of Self-Attachment. Finally, we design a cross-platform application as the bot's user interface. We evaluate our platform in a ten-day human study with N=52 volunteers from the non-clinical population, who have had over 2,000 dialogues in total with the chatbot. The results indicate that the platform was engaging to most users (75%), 72% felt better after the interactions, and 74% were satisfied with the SAT Teacher's performance.

Figures (4)

• Figure 1: The pipeline of the SAT chatbot.
• Figure 2: The conversation flowchart of the chatbot, adapted from 9750315.
• Figure 3: User interface of our platform in Chatbot and SAT Teacher modality.
• Figure 4: The chatbot's evaluation results.