skip to main content
10.1145/3698587.3701489acmconferencesArticle/Chapter ViewAbstractPublication PagesbcbConference Proceedingsconference-collections
abstract

Building Personality-Adaptive Conversational AI for Mental Health Therapy

Published: 16 December 2024 Publication History

Abstract

Many people with mental health problems cannot get professional help for various reasons such as lack of awareness, unavailability, unaffordability, etc. A virtual conversational agent can offer an alternative to deliver mental health care that is accessible, affordable, and scalable. However, building such agents using a one-size-fits-all approach may not be effective for everyone, as different individuals have different personality types that dictate how they communicate with chatbots. Therefore, developing therapy chatbots that can adjust to the user's personality is important. In this work, we present the important role of personality-adaptive conversational agents (PACAs) in the context of mental healthcare. We designed an architecture around traditional machine learning (ML) models and open-source large language models (LLMs) to build a PACA for mental health therapy, developed a working prototype based on it, and conducted a user study to conclude that personality-adaptiveness is indeed an important feature for mental health chatbots.
Our research was based on the iCare Project [1], and the associated development was meant to minimize the limitations of the project. We designed the PACA to adapt its responses according to the personality profile of the user created over time. The personality profiles were based on the results from a text classification model fine-tuned for the Big Five Personality Traits[2] with a classification accuracy of 96%. We self-hosted an open-source LLM which made use of accumulated personality information through prompt engineering. The final setup was able to generate adapted therapeutic responses with an average response time of 10 seconds using different hyperparameter tuning and fine-tuning approaches.
We conducted a user study among 20 subjects to compare the performance of the personality-adaptive chatbot with its non-adaptive counterpart and found that the adaptive feature almost doubled the number of users who found the chatbot relevant and helpful for their mental healthcare needs. 75% of users felt comfortable discussing any sensitive topic with the adaptive chatbot in comparison to 45% for the non-adaptive chatbot. The PACA prototype validated the feasibility of creating accessible personalized mental healthcare solutions using advanced ML techniques and the results from the user study highly recommend the implementation of PACA into mental health chatbots. The prototype is currently live and freely available for use.

References

[1]
Remya Mavila, Sugam Jaiswal, Raghav Naswa, Weichao Yuwen, Bill Erdly, and Dong Si. 2024. iCare - An AI - Powered Virtual Assistant for Mental Health. In 2024 IEEE 12th International Conference on Healthcare Informatics (ICHI). 466--471. https://doi.org/10.1109/ICHI61247.2024.00066
[2]
Sonia Roccas, Lilach Sagiv, Shalom H Schwartz, and Ariel Knafo. 2002. The big five personality factors and personal values. Personality and social psychology bulletin 28, 6 (2002), 789--801.

Recommendations

Comments

Information & Contributors

Information

Published In

cover image ACM Conferences
BCB '24: Proceedings of the 15th ACM International Conference on Bioinformatics, Computational Biology and Health Informatics
November 2024
614 pages
ISBN:9798400713026
DOI:10.1145/3698587
Permission to make digital or hard copies of part or all of this work for personal or classroom use is granted without fee provided that copies are not made or distributed for profit or commercial advantage and that copies bear this notice and the full citation on the first page. Copyrights for third-party components of this work must be honored. For all other uses, contact the Owner/Author.

Sponsors

Publisher

Association for Computing Machinery

New York, NY, United States

Publication History

Published: 16 December 2024

Check for updates

Author Tags

  1. chatbot
  2. conversational AI
  3. fine-tuning
  4. large language models
  5. mental healthcare
  6. paca
  7. personality
  8. prompt engineering

Qualifiers

  • Abstract
  • Research
  • Refereed limited

Conference

BCB '24
Sponsor:

Acceptance Rates

Overall Acceptance Rate 254 of 885 submissions, 29%

Contributors

Other Metrics

Bibliometrics & Citations

Bibliometrics

Article Metrics

  • 0
    Total Citations
  • 114
    Total Downloads
  • Downloads (Last 12 months)114
  • Downloads (Last 6 weeks)46
Reflects downloads up to 19 Feb 2025

Other Metrics

Citations

View Options

Login options

View options

PDF

View or Download as a PDF file.

PDF

eReader

View online with eReader.

eReader

Figures

Tables

Media

Share

Share

Share this Publication link

Share on social media