Dr. Chengxiang Zhai

Donald Biggar Willett Professor in Engineering, ACM SIGIR Gerard Salton Award Winner (2021), University of Illinois Urbana-Champaign

Title: The Future of Digital Libraries: Serving Humans and Artificial Intelligence Agents

Abstract: Digital libraries have traditionally been designed to serve human information consumers through information retrieval (IR) systems. The rapid emergence of large language models and increasingly capable artificial intelligence agents is fundamentally changing this landscape. First, agentic AI technology enables library IR systems to go beyond information access to more actively provide task support for users in a personalized manner, transforming passive library IR systems into active, personalized, intelligent information-serving environments. Second, as AI agents become increasingly capable of searching, reading, reasoning over, and acting upon vast collections of information on behalf of people, future digital libraries will need to serve not only humans, but also AI agents. Motivated by these observations, in this talk, I will explore a vision of the next generation of digital libraries as trustworthy, personalized, intelligent information environments that support both human and machine users. I will review some recent progress toward this vision and discuss the major challenges that remain. I will argue that with the rapid growth of AI-generated content, information quality will become a major and increasingly difficult challenge in the future, potentially undermining the trustworthiness of all AI systems and that digital libraries hold the key to addressing this critical challenge. I will conclude with a vision of digital libraries evolving into a broader interdisciplinary research community in the future at the intersection of information science, computer science, artificial intelligence, and the social sciences.

Bio: ChengXiang Zhai is a Donald Biggar Willett Professor in Engineering in the Siebel School of Computing and Data Science at the University of Illinois Urbana-Champaign, where he also holds joint appointments in the School of Information Sciences, the Carl R. Woese Institute for Genomic Biology, and the Department of Statistics. He earned a Ph.D. in Computer Science from Nanjing University in 1990 and a Ph.D. in Language and Information Technologies from Carnegie Mellon University in 2002. From 1997 to 2000, he worked at Clairvoyance Corp., first as a Research Scientist and later as a Senior Research Scientist. His research interests are broadly in intelligent information systems, with a particular focus on intelligent agents that augment human intelligence through interactive information retrieval and data analytics. His research aims to develop general frameworks, models, and algorithms for interactive information retrieval and data analytics and apply them to healthcare, education, and accelerating scientific discovery. He has published more than 400 peer-reviewed papers, with more than 45,000 citations and an h-index of 98 according to Google Scholar. He holds five U.S. patents. He is America Editor of the Springer Information Retrieval Book Series. He has served as an Associate Editor for multiple journals (e.g., ACM TIST, ACM TOIS, IPM, and ACM TKDD), as Program Co-Chair for ACM CIKM’04, NAACL HLT’07, ACM SIGIR’09, ECIR’14, and WWW’15, and as General Co-Chair for CIKM’16, WSDM’18, and IEEE BigData’20. He is an ACM Fellow and a member of the ACM SIGIR Academy. His honors include the ACM SIGIR Gerard Salton Award, three ACM SIGIR Test of Time Awards, the 2004 Presidential Early Career Award for Scientists and Engineers (PECASE), an Alfred P. Sloan Research Fellowship, an IBM Faculty Award, and an HP Innovation Research Award.

Dr. Nicholas Belkin

Distinguished Professor Emeritus of Library and Information Science, ACM SIGIR Gerard Salton Award Winner (2015), Rutgers University

Title: Digital Libraries and Their Technologies: A Look Back and a Peek Forward

Abstract: The basic ideal underlying digital libraries; that is, of universal, remote, access to the contents of the libraries of the world, has long been a goal of librarians and information scientists, among distinguished others. But the dream only began to become realized when the required technologies began to become available. And, as the relevant technologies have progressed, the field of digital libraries has responded (or, one might say, reacted) accordingly. Until quite recently, this response has been pretty much as one would expect; expansion of the materials in the libraries, increased and better methods of access to their contents, and better interaction between the libraries and their users. But with the appearance of Generative LLMs, there now appears to be some question as to whether digital libraries can continue in this same path. In this talk, I look back at the beginnings of the current digital library movement and how digital libraries have responded to technological advances, and consider some possibilities for their future in the face of ubiquitous general access to Generative LLMs.

Bio: Dr. Nicholas J. Belkin is Distinguished Professor of Information Science Emeritus at Rutgers University and Adjunct Professor at the Dhirubhai Ambani Institute of Information and Communication Technology. He is widely recognized for his contributions to information science and interactive information retrieval, especially as one of the founders of the “cognitive viewpoint” in information science. He co-authored one of the first books to examine interactive information retrieval as a process, and has published more than 200 journal articles, conference papers, and book chapters. His research has addressed information behavior, personalization, search interaction, and whole-session search evaluation, including twelve years of work in the TREC Interactive Track. Dr. Belkin has served as Chair of ACM SIGIR and President of ASIS&T, and received the ASIS&T Award of Merit and ACM SIGIR Gerard Salton Award. He also chaired the organizing committee that combined ACM and IEEE-CS Digital Library conferences into the one JCDL.

Dr. Leif Azzopardi

Title: Design for Information Farmers: Digital Libraries as Seeds for Cultivating Knowledge

Abstract: For decades, digital libraries have focused on helping users discover, access, and assemble information from large collections. Yet emerging generative and agentic technologies are changing how people satisfy their information needs. Recent work introduced the concept of Information Farming to describe this shift. Rather than finding information that exists “out there,” users are increasingly cultivating information through their interaction with generative systems. If users are becoming farmers rather than foragers, then what does this mean for Digital Libraries?

This talk explores the implications of Information Farming for the future of Digital Libraries. I argue that digital libraries need to evolve from repositories of information artifacts into seeds for cultivating knowledge: infrastructures that preserve, structure, and expose knowledge in forms that can be cultivated by human/agent teams. In such a future, the value of a digital library may lie not only in what it contains, but in what can be grown from it.

This perspective raises a multitude of questions. What should digital libraries preserve when documents are no longer the primary unit of interaction? How should provenance and trust be maintained when information is increasingly farmed rather than foraged? Should digital libraries provide content for agents to consume, or become generative environments in their own right? I do not argue that foraging will disappear. Rather, I argue that Digital Libraries must be designed for a world in which foraging is no longer the dominant mode of knowledge work.

Bio: Dr. Leif Azzopardi has pioneered methods for modelling and measuring how information retrieval systems shape users’ behaviour and decisions. Over the past 20 years, his work has centred on three themes: user simulation for evaluating systems, bias and fairness in search engines, and the economics of search. This has led to notable contributions including using language models to generate simulated queries for test collections, years before synthetic query generation became common; introducing the concept of retrievability to measure how retrieval biases make some documents more likely to be seen than others; and applying economic theory to explain search behaviour, and show how users subscribe to the principle of Least Effort in Finding. More recently, he has questioned the influence and impact of cognitive bias in search, explored whether personalised search can support ethical consumerism, and reconsidered how generative AI changes information interaction. He has posited that we are moving from berry picking and information foraging toward berry growing and information farming. Dr. Azzopardi is a Principal Applied Scientist at Microsoft, where he leads measurement and evaluation of enterprise search quality within Copilot, and an Associate Professor in the Department of Computer and Information Sciences at the University of Strathclyde.