CIIR Talk Series: Souvick Ghosh, Human-Centered AI for Information Seeking and Decision-Making

Content

Speaker: Souvick Ghosh (San José State University)

Image

Image of Dr. Souvick Ghosh

Abstract: AI now mediates how people search, sense-make, and decide, across everyday queries, high-stakes judgments, and institutional workflows. In this talk, Dr. Ghosh outlines a human-centered agenda for ethical AI that treats conversation as a primary interface for information seeking and decision support. He will discuss his ongoing work on ethical AI that surfaces bias and uncertainty, makes model behavior intelligible, minimizes interaction burden, and respects cultural and legal contexts. Drawing on collaborations with libraries, archives, and community partners, Dr. Ghosh translates these ideas into practical patterns for search and chat systems. The goal is to develop AI systems that not only perform well but also explain themselves, invite scrutiny, and improve community well-being.

Bio: Dr. Souvick Ghosh is an Assistant Professor at the School of Information, San José State University. His work sits at the intersection of information retrieval, human–computer interaction, and responsible AI. He designs and studies conversational and decision-support systems that foreground accountability, explainability, and cultural awareness. Specifically, his focus is on improving the lives of marginalized communities through the application of trustworthy AI. Partnering with libraries, archives, and civic organizations, he develops methods and interfaces that help people interrogate algorithmic outputs, reason under uncertainty, and access information equitably. He integrates this research into teaching and outreach on responsible computing, preparing practitioners and communities to build AI systems that are transparent, auditable, and centered on human needs. The efforts of Dr. Ghosh and his colleagues were mentioned in the White House fact sheet.