My research develops novel computational methods and technologies powered by these methods to employ social media, responsibly and ethically, toward improving our health and well-being.

As social media technologies are adopted more pervasively, the line between our online and offline lives is disappearing slowly but steadily. Content shared on these platforms often revolves around day-to-day happenings and experiences in our personal lives, and in our social environments. As such, social media provides a means to capture attributes relevant to our thinking, mood, communication, activities, socialization, and psychological states. How can we employ this rich resource in addressing outstanding challenges relating to personal and societal well-being?

Among different well-being concerns, mental and psychological disorders like mood disorders can have wide-reaching effects on peoples daily activities, education, employment, occupational functioning, and relationships. They are a leading cause of disability worldwide. According to the National Institutes of Mental Health, one in four adults, or 61.5 million Americans, are reported to experience such a challenge in a given year. However currently these well-being concerns are primarily assessed through self-report based mechanisms, with limited reliance on evidence-based assessment methods. These methods require retrospective recollection of somewhat subjective wellbeing facts, and are therefore vulnerable to patients' memory bias. Time and budgetary constraints often limit practicing clinicians from conducting more thorough and frequent diagnostic evaluations. Moreover, the social stigma associated with mental health precludes people from seeking the care they need. Consequently, the need for proactive detection of mental illness is emphasized, including alternative mechanisms to extend timely support to vulnerable individuals.

Even with timely detection and diagnosis, treatment and management of mental illnesses remain difficult. It is established that medications and therapy alone may be insufficient for managing these illnesses -- many individuals experience persistent symptoms even with ongoing pharmacological care. Importantly, a commonality of many mental health challenges is their high likelihood of relapse. Consequently, aside from traditional treatment, self-care and social support based interventions have been recognized to help meet some of these challenges. Designing, developing, and deployment appropriate technological interventions remain an active area of research.

Many mental well-being challenges are known to be characterized by latent processes that include negative perspectives, self-focused attention, loss of self-worth and self-esteem, and social disengagement. Today, with social media, many of these latent processes, such as the context and content of such as ones affective, behavioral, and cognitive reactions can be observed in real-time, longitudinally, and unobtrusively. At the same time, individuals with well-being challenges are increasingly appropriating social media platforms to engage in candid disclosures of various well-being challenges. This has led to the emergence of many online communities that aim to provide scaffolding to individuals at risk. Taken together, my research agenda examines two broad questions: whether we can employ passively shared data on social media to pro-actively detect ones risk to mental well-being challenges, and second, how social media platforms can facilitate interventions for improving one's well-being.

Towards these two goals, my research program adopts a highly interdisciplinary and collaborative approach, that carefully balances methodological contributions with practical impact. I combine the power of social computing, machine learning, and natural language analysis with insights and theories from the social, behavioral, and clinical science fields, such as psychology, sociology, psychiatry, and public health. I evaluate the developed computational approaches using diverse human-centered methods, deriving conclusions grounded in human behavior.

For further information on our specific projects and publications, please check out the Publications page.