My research helps people find, understand, and create text information. I apply a combination of information retrieval, data/text mining, and HCI methods to study search engines, conversational agents, writing assistants, and other text information systems.
Dr. Jiepu Jiang is an Assistant Professor in the Information School of the University of Wisconsin-Madison, which is a part of the School of Computer, Data & Information Sciences (CDIS). Before joining UW-Madison, he was an Assistant Professor of Computer Science at Virginia Tech from 2018 to 2020. His research interests lie at the intersection of information retrieval and HCI, especially studying human-AI interaction issues in various text information systems. He holds bachelor's and master's degrees from Wuhan University, a Ph.D. in Library & Information Science from the University of Pittsburgh, and is finishing another Ph.D. in Computer Science at the University of Massachusetts Amherst.
Conversational Agents & Chatbots
I study conversational agents, both text and voice-based ones. My recent research focuses on human-AI hybrid chat systems, using IR or NLG-based text suggestions to help people finish knowledge-demanding conversations. I have also worked on user satisfaction prediction in intelligent assistants.
Interactive Information Retrieval
I study human factors and behavior in search systems, including topics such as search task, search session, query reformulation, click decision, eye and mouse movement, relevance judgment, intelligent user interface, and misinformation and bias in web search results.
Search Quality & User Experience
I study algorithmic and data-centric methods for evaluating search quality and user experience, e.g., adaptive evaluation measures that better align with user experience than conventional ones such as NDCG and user satisfaction predictive models based on behavioral signals.
Explainability of Search Interfaces
I study search system's explainability, especially how search interface informs and explains search systems to users, e.g., how search result summary's explainability affects click, and how to improve search result summaries to explain relevance and non-relevance better.