Send message to Authors

Short Abstract

We share reflections on an AI-enhanced redesign plan of a mental health citizen science chatbot, leveraging its rich dataset to build a digital participatory tool that serves as an individualized resource, offering inclusive support, fostering adaptive dialogue, and amplifying community voices.

Abstract

CoAct for Mental Health is a citizen social science initiative that individuals with mental health lived experiences and their families as co-researchers. They documented their personal experiences on mental health social support networks through micro-stories. A Telegram chatbot then disseminated these micro-stories, inviting subscribers to reflect on whether they, and someone around them, have ever lived similar experiences. More than 11,000 responses from more than 400 participants were recorded, generating a rich dataset that also includes sociodemographic profiles and indicators of emotional well-being.

We propose to study this dataset through the lens of social support networks, using AI and network-based modelling from a complex systems perspective. This analysis uncovers patterns of participation and identifies key lived experiences linked to specific sociodemographic profiles, revealing how mental health experiences emerge and can be strengthened within communities. This approach gives value to the voices shared, transforming them into actionable knowledge to enhance mental health services.

Building on these insights, it is possible to reflect on how AI can help the chatbot to adapt dynamically to participants. In this way, we want to sustain engagement more effectively while fostering self-reflection and deliver more informative data to keep studying social support networks in a participatory and open manner.

Ultimately, the envisaged AI-driven redesign would like to offer more inclusive support, empowering citizens, and providing a data-informed foundation for improving mental health care and policy. Therefore, this participatory platform bridges the gap between social needs expressed by communities and institutional policymaking.