Recently, a paper entitled “Dynamic Recommendation of Physician Assortment with Patient Preference Learning” published on IEEE Transactions on Automation Science and Engineering (IEEE T-ASE) won the prestigious 2020 IEEE T-ASE Best Paper Award. The authors are Dr. Xin Pan who graduated from the College of Engineering, Dr. Jie Song, Associate Professor, and Ph.D. student Fan Zhang from the Department of Industrial Engineering and Management, College of Engineering. The award, which recognizes the best paper of the IEEE T-ASE published in the year of 2019, has been announced and presented at the annual 2020 IEEE Conference on Robotics and Automation (ICRA) on June 6th, 2020.
IEEE T-ASE is a renowned scientific journal in the international intelligent control and optimization community, published by IEEE Robotics and Automation Society (IEEE RAS), the largest international society of robotics and automation in the world. IEEE T-ASE, now in its sixteenth year, aims to award research papers with outstanding theoretical results and significant technical value in the field of intelligent control and optimization.
Motivated by a popular physician recommendation application on a web-based appointment system in China, this award-winning paper gives a pioneer work in modeling and solving the physician recommendation problem with heterogeneous illness conditions of patients by using optimization theory, machine learning technology and statistical inference method. In order to solve the challenge brought by the lack of data on patient preferences, the preference learning algorithm is proposed that optimizes the recommendations and learns patient preferences at the same time. The online learning of patients preference is realized by automatically collecting patient feedbacks to recommendations of physician and continuously updating the choice model of patients. Under the goal of optimizing the recommendation of physician, personalized optimal recommendations of physician are provided.
The research results not only help to improve the experience of patients in the web-based appointment system and the use of high-quality physician resources, but also can be extended to other types of web-based service platforms and improve the corresponding users’ experience and the use of high-quality service resources.
The IEEE RAS committee highly appraised the theoretical value and wide application prospect of this paper. According to the research results, Prof. Jie Song’s group has cooperated with the leading web-based appointment platform HaoDF Online, aiming at providing the optimization of supply and demand matching strategy between physicians and patients online, optimizing physician resource allocation, providing efficiency management and operation of the platform, and being committed to the development and improvement of internet medical services.