Real-time bidder feedback in continuous combinatorial auctions
A novel computational framework capable of providing real-time bidder support for combinatorial auctions.
Applications
- Online auctions
- Large-scale asset sales
- Procurement and reverse auctions
Key Benefits & Differentiators
- Real-time feedback: Enables bidders to receive instant information on bid viability and competitivenes
- Optimized bidder decisions: Computes critical metrics such as bid winning probability and deadness levels
- Scalable: Implements data structures and algorithms for practical auction sizes
Technology Overview
Combinatorial auctions, where bidders can place bids on individual items or item bundles, are crucial for many industries but remain difficult to implement in general-purpose marketplaces. A major challenge is the lack of real-time bidder support, which limits auction transparency and bidder strategy optimization. Traditional single-item auctions, such as those on platforms like eBay, provide immediate feedback, but multi-item multi-unit (MIMU) and single-item multi-unit (SIMU) auctions require significantly more complex computations to evaluate bids dynamically. This complexity has hindered the adoption of combinatorial auctions in broader business and consumer markets.
Researchers at the University of Minnesota have developed computational infrastructures capable of providing real-time bidder support for combinatorial auctions of practically relevant sizes. This framework provides real-time feedback to bidders in continuous combinatorial auctions, where participants can join and leave the auction at any time. The approach accommodates both OR bidding and XOR bidding, ensuring flexibility in bidding preferences. To optimize bidder decision-making, the technology breaks down an auction into sub-auctions, allowing bidders to assess their best strategies based on known bids. Additionally, this technology can be extended to combinatorial reverse auctions, a commonly used mechanism for industry procurements.
Phase of Development
TRL: 4-5Prototype; algorithm developed (python code w/ simulation).
Desired Partnerships
This technology is now available for:- License
- Sponsored research
- Co-development
Please contact our office to share your business’ needs and learn more.
Researchers
- Alok Gupta, PhD Curtis L. Carlson Chair in Information Management, Information & Decision Sciences, Carlson School of Management
- Gediminas Adomavicius, PhD Professor, Information & Decision Sciences, Carlson School of Management
- Mochen Yang, PhD Associate Professor, Information & Decision Sciences, Carlson School of Management
-
expand_more library_books References (2)
- Gediminas Adomavicius, Alok Gupta, Mochen Yang (2022), Bidder Support in Multi-item Multi-unit Continuous Combinatorial Auctions: A Unifying Theoretical Framework, Information Systems Research, 33, 1174-1195
- Gediminas Adomavicius, Alok Gupta, and Mochen Yang (2019), Designing Real-Time Feedback for Bidders in Homogeneous-Item Continuous Combinatorial Auctions, MIS Quarterly, 43, 721-743
-
expand_more cloud_download Supporting documents (1)Product brochureReal-time bidder feedback in continuous combinatorial auctions.pdf