Zamani Receives NSF CAREER Award for Work on Conversational Information Retrieval
Manning College of Information and Computer Sciences (CICS) Assistant Professor Hamed Zamani received a CAREER Award totaling more than $570,000 from the National Science Foundation (NSF) for the expansion of his project, Enriching Conversational Information Retrieval via Mixed-Initiative Interactions.
While the importance of natural language conversations to the future of efficient and effective conversational search engines and technology has become widely accepted, Zamani’s research suggests that existing systems cannot support optimal natural conversational interactions with users in their current iteration.
Traditional search systems are designed based on a query-response paradigm, in which the user initiates a search, and the system responds by generating one or more results that its algorithm has deemed relevant to the user’s search request. However, users may find themselves repeating searches or, ultimately, giving up after receiving results that fail to address the intention behind user-generated search requests.
“This often time-consuming process repeats until the user terminates the search session and frequently doesn’t give the user the answer they are looking for,” describes Zamani. “My work explores ways to make conversational search more effective through the use of mixed-initiative interactions, where, for example, the system may ask a clarifying question instead of providing an answer to ambiguous questions or recommend new information even though it is not an explicit response to the search query.”
Zamani’s CAREER project addresses a key aspect of the future of search technology by providing access to information through natural language conversations. It aims to advance the state-of-the-art in conversational search by envisioning solutions that consider mixed-initiative interactions by studying theoretical foundations for measuring mixed-initiative conversations; models for clarifying the user's information needs; and models for proactive informational contributions to ongoing conversations.
“This project addresses crucial problems in spoken conversational systems that can be widely used by millions of visually impaired users and everyone else who is visually occupied, such as car drivers,” explains Zamani. “In addition, since dialogue is the most natural form of human communication for exchanging knowledge, advances in conversational information retrieval have broad applications for children, especially for educational purposes.”
Zamani’s project also plans to develop several resources for advancing the field of conversational information retrieval, including a conversational scholarly assistant agent to be used as a tool for online experimentation and public data creation.
Hamed Zamani is the Associate Director of the Center for Intelligent Information Retrieval (CIIR). He joined CICS in September 2020. Prior to UMass Amherst, he was a researcher at Microsoft, working on a wide range of problems related to search engines. He received his doctorate in computer science from UMass Amherst in 2019 and was a recipient of the UMass Amherst CICS Outstanding Dissertation Award for his PhD thesis on weakly-supervised neural information retrieval.
The CAREER Award is one of NSF's most prestigious awards in support of junior faculty who exemplify the role of teacher-scholars through outstanding research, excellent education, and the integration of education and research, leading to advances in the mission of their department or organization.