Building Better Online Support Groups: A Smart Approach to Mental Health Communities on Social Media

Mental health challenges are a growing concern worldwide, with many individuals turning to social media for support. Online peer support groups offer a space for people to connect, share experiences, and find encouragement. However, most existing groups form organically or are created by healthcare professionals without a structured approach. This can result in groups that are not always the most effective for those in need. Our research aims to address this gap by developing a method to systematically create strong, well-connected support groups on social media. Instead of relying on chance, our approach identifies the most suitable members for a support network based on shared interests and social proximity. By improving the way these groups are formed, we can enhance the effectiveness of online peer support and provide better mental health resources for individuals in need.

To build more effective support groups, we designed an algorithm that finds people on social media who are socially close to a target user and share similar interests. This ensures that individuals are placed in groups where they are more likely to receive meaningful support.

One of the biggest challenges was optimizing this group formation process, as finding the best possible support network is a computationally complex problem. Our research proves that this problem is difficult to solve optimally, but we developed an efficient algorithm that produces high-quality support groups quickly. Using real-world data from Meta, we demonstrated that our approach effectively forms well-connected groups very efficiently. This allows for the rapid creation of structured support networks that can be scaled for larger populations.

The way support groups are formed can significantly impact mental health outcomes. Our approach ensures that individuals struggling with depression or other mental health conditions are placed in the best possible groups, helping reduce feelings of loneliness and isolation.

For healthcare organizations and social media platforms, this research provides a practical way to improve online peer support. Instead of leaving group formation to chance, our method enables the creation of structured, data-driven support networks. This can lead to better mental health interventions, encourage more people to seek help, and create safer, more supportive online spaces. From a broader perspective, this work aligns with the growing use of AI and social computing for mental health solutions. By leveraging technology, we can build more effective and scalable support systems, making mental health care more accessible to those who need it most.

Social media has the potential to be a powerful tool for mental health support, but its effectiveness depends on how connections are made. Our research offers a new way to systematically form support groups that are better suited to helping individuals cope with mental health challenges. By using AI and data-driven methods, we can create online communities that provide stronger emotional support, reduce isolation, and improve mental well-being. As we continue refining this approach, we hope it can be implemented on a larger scale to support mental health initiatives worldwide.Yang, X., Li, G., & Noorian, Z. (2025). Support group formation for users with depression in social networks (external link, opens in new window) . Expert Systems with Applications, volume 276, 127107. DOI: 10.1016/j.eswa.2025.127107