Dongsheng Li is currently working with the Sales Science team in IBM Research - China. Before joining IBM, he worked as a postdoc research at the department of computer science and technology, Tongji University (supervised by Prof. Li Shang). He obtained Ph.D. from school of computer science, Fudan University, in 2012 (supervised by Prof. Ning Gu), and obtained B.E. from department of computer science and technology, University of Science and Technology of China (UTSC), in 2007. He also visited University of Colorado Boulder as a visiting scholar from 2010.8 to 2011.2 (supervised by Prof. Qin Lv).
He is a member of China Computer Federation (CCF). He served as PC members for AAAI 2017 and AAAI 2018. He also served as reviewers for many well-known conferences/journals, e.g., Neurocomp, KBS, FGCS, IJCIS, IEEE Access, NIPS, IJCAI, AAAI, etc.
Dongsheng Li's research interests focus on key issues in recommender systems, including but not limited to accuracy, efficiency, stability, scalability, privacy and security of collaborative filtering algorithms. Besides, he is also interested to general machine learning applications, including matrix approximation methods, algorithm stability, generalization performance analysis, noise-resilience analysis and graphical models, etc. Meanwhile, he has also been working on data-driven sale science topics with IBM, including up sale, cross sale, opportunity discovery, etc.
Besides computer science, he is also interested to data analysis in smart grid applications, e.g., user appliance usage analysis in smart grid, privacy-preserving smart metering framework design and data-driven fault prognosis and diagnosis in wind turbine and PV systems.