How can we better understand people through the language they use, and how can AI systems learn to engage with us in more socially aware and responsible ways? The Computing for Social Good (CSG) Lab at Northeastern University’s Khoury College of Computer Sciences explores these questions by studying language, behavior, and interaction in real world contexts.

Our research centers on large language models and their role in society. We investigate how LLMs perceive and respond to emotional and social signals, and how these systems can be aligned to act safely, responsibly, and in ways that reflect human values. In parallel, we examine socially grounded language phenomena, including hate speech, empathy, emotion, and mental health, with the aim of understanding how these dynamics emerge in both human communication and AI generated text.

The CSG Lab takes an interdisciplinary approach that connects natural language processing, computational social science, and machine learning with perspectives from fields such as psychology and the social sciences. We develop methods that not only improve model performance, but also shed light on how these systems reason about people, relationships, and society.

Beyond advancing theory, we are committed to building practical tools, datasets, and evaluation frameworks that support the broader research community. Our goal is to contribute to AI systems that are not only more capable, but also more socially aware, transparent, and aligned with human well-being.

We collaborate with researchers, students, and partners across disciplines who share an interest in using AI to address meaningful societal challenges.