Ongoing Research Projects
Currently, Di is actively working with his PhD and Master’s students on the following topics, in collaboration with a few industrial partners and other academic institutes (mainly including Tencent, Wedge Networks, Meridian Lightweight Technologies, and University of Toronto):
- Text modeling, text understanding, text matching, information retrieval, search, event discovery and summarization, natural language generation and concept extraction.
- Distributed machine learning and data analytics processing algorithms and platforms, with a focus on both algorithm design and innovative system development. The specific problems we study include distributed algorithms for training machine learning models, such as deep neural networks and matrix factorization, machine learning based on decentralized data for privacy preservation and efficiency optimization.
- Auto machine learning, active learning, data generation, active data selection, and data/model quality assessment.
- Spatial-temporal data mining for social-economic computing.