Several YWCC Researchers Contribute to KDD 2024

Distinguished Professor Guiling (Grace) Wang in the Ying Wu College of Computing (YWCC) recently co-organized a special event, Finance Day, under the theme of “Finance” as part of KDD 2024 held in Barcelona Spain with her collaborator, Daniel Barraio, AI research director at JP Morgan Chase. Professor Baruch Schieber and Associate Professor Senjuti Basu Roy, along with MD Mouinul Ph.D. ’23, and Ph.D. candidate Soroush Vahidi, also published a paper at the research track of the conference titled Promoting Fairness and Priority in k-Winners Selection Using IRV. All are from the Department of Computer Science.

KDD Finance Day, initiated by Wang from last year, brings together experts in Data Mining, AI, Economy, and Finance for fruitful discussions, fostering collaboration between academia and industry, and delves into the latest developments in FinTech technology and its promising future. This year’s event featured 12 talks by invited speakers from eight different countries, including US, UK, Germany, Serbia, Norway, Netherlands, South Korea and Hong Kong China.

Wang was recognized by KDD for her success in coordinating not only the event itself but arranging for the inclusion of distinguished featured speaker Henrike Mueller, AI strategy manager, Financial Conduct Authority, UK, and keynote speaker Avanidhar Subrahmanyam, distinguished professor, Anderson School of Management, UCLA, and Goldyne and Irwin Hearsh chair in Money and Banking.

The published paper by Schieber, Roy, Vahidi and Mouinul, who is currently a senior machine learning research scientist at PayPal, studies applicability and computational implications of adapting Instant Run-off Voting or IRV to repeatedly select one winning candidate given a very large number of incomplete preference orders over many candidates, while satisfying group fairness and a priority order. IRV (Instant Run-off Voting) is a ranked choice voting mechanism that has been gaining popularity lately as an electoral system in Australia, Ireland, and the U.S because it promotes proportional representation of solid coalition and anti-plurality, as well as reduces strategic voting. The work demonstrates the suitability of the proposed models and designed solutions in many compelling applications that benefit from fair decision making, such as, hiring candidate(s) for a job, selecting member(s) of a committee, finding winning candidate(s) in a competition, in electoral voting, or even in recommender systems.