Transforming Mental Health Care with AI and Language TechnologySpeaker: Tanmoy Chakraborty – New Delhi, India
Topic(s): Artificial Intelligence, Machine Learning, Computer Vision, Natural language processing
Mental health issues are rising at an alarming rate. A recent report reveals that one out of six people suffers from mental health related issues. At the same time, there is a severe shortage of mental health experts to facilitate adequate support to clients. Earlier research shows that at least every one in five youths carries mental health issues. Nine out of ten live in less developed countries, barely having any access to advanced healthcare facilities. The overburdened healthcare system has necessitated the development of alternative solutions to address people's treatment needs, such as Virtual Mental Health Assistants (VMHAs). For instance, popular chatbots such as Woebot and Wysa are known for assisting help-seekers. Similarly, peer counseling platforms like Reddit, TalkLife, and 7Cups allow immediate therapy appointments worldwide. Although advancements in this area may seem promising, the foundational aspects of VMHAs still encounter obstacles. Our research focuses on (a) counseling dialogue understanding, (b) rich dialogue generation in VMHAs, (c) understanding support-seekers in various counseling methods, and (d) counseling summarization.
Understanding users' characteristics and intents is crucial to monitor their mental health status. However, with the advent of large language models, the essence of intents may be diminished. Nonetheless, dialogue-acts are essential attributes that need to be comprehended in any conversational system. They play a crucial role in shaping the direction of a conversation and determining the next response from the agent. To ensure a well-functioning VMHA, researchers have explored specific systems that predict dialogue-acts in precise counseling setups. In addition, generating counseling summaries is also essential for addressing revisiting clients on the bot. The challenge in counseling summarization lies in incorporating essential domain knowledge. Our work focuses on generating dialogue summaries that pertain to counseling conversations and addressing these challenges by exploiting domain knowledge. By addressing these challenges of understanding dialogue-acts and generating counseling summaries, we contribute towards a well-functioning VMHA that provides an end-to-end counseling experience to its users. Our work has implications for the future development of conversational agents in the domain of Mental Health Counseling, and we believe it will help advance the field toward more effective and accessible mental health care.
About this LectureNumber of Slides: 50
Duration: 45 minutes
Languages Available: English
Request this Lecture
To request this particular lecture, please complete this online form.
Request a Tour
To request a tour with this speaker, please complete this online form.
All requests will be sent to ACM headquarters for review.