This inventor holds 1 USPTO granted patent and 1 published patent application. Active years: 2026.
Data source: USPTO (United States Patent and Trademark Office) public patent records. Weekly synchronization. How IDiyas builds this profile
Years Active: 2026
Title: Kun Jin and Fairness Measurement in Recommendation Systems.
Introduction
Kun Jin is an inventor based in Culver City, California. His patent work addresses how fairness can be measured in large-scale recommendation systems when some labels are missing.
Latest Patents
Jin is associated with 1 patent, titled “Measuring Fairness in Large-Scale Recommendation Systems with Missing Labels.” The patent describes computer-implemented methods and systems for estimating fairness metrics. For each user, the method obtains data about items recommended by a recommendation model and data about a separate collection of items selected randomly for recommendation. It then calculates a fairness metric using both sets of data, offering an approach to fairness measurement when labels are incomplete.
Career Highlights
The patent highlights Jin’s work on evaluating recommendation systems, particularly the challenge of assessing fairness when available data does not provide complete labels. Its method combines model-generated recommendations with randomly selected recommendations as inputs to the fairness calculation.
Collaborations
Jin’s listed coworkers include Yulong Dong and Xinghai Hu.
Conclusion
Kun Jin’s patent focuses on a practical method for estimating fairness in large-scale recommendation systems despite missing labels.
Data source: USPTO (United States Patent and Trademark Office) public patent records. Weekly synchronization. How IDiyas builds this profile