Room 1421, Building L, Harbin Institute of Technology (Shenzhen), Shenzhen, China  PC, Postcode: 518055



(0755) 2690 5221

Selected Projects

  • University Ranking

    University ranking endeavors to play an unimaginable role in motivating and recognizing excellence, which greatly facilitates a wide range of stakeholders. Regarding such ranking, criticism often co-exists with praise. The controversial points of the current university ranking systems are threefold: 1) insufficient data, 2) labor-intensive user survey, and 3) naive fusion of multi-channel data. Towards these end, we present in this project a novel university ranking scheme to automatically and transparently rank universities by harvesting large-scale Web data from various views. The ranking results of Chinese universities are available here:

  • Micro-video Analysis

    The unprecedented growth of portable devices contributes to the success of micro-video sharing platforms such as Vine, Kuaishou, and Tik Tok. They enable users to record and share their daily life within a few seconds in the form of micro-videos at any time and any place. As a new media type, micro-videos gain tremendous user enthusiasm, in virtue of their value in brevity, authenticity, communicability, and low-cost. Yet, micro-videos have their unique characteristics and the corresponding research challenges, including but not limited to the followings: 1) Information sparseness. 2) Hierarchical structure. 3) Low quality. 4) Multimodal sequential data. And 5) no public benchmark dataset. To tackle the aforementioned research challenges, we present some state-of-the-art multimodal learning theories and verify them over three practical tasks of micro-video understanding: popularity prediction, venue category estimation, and micro-video routing. You can find the packed codes and data here:

    You can enjoy this book via

  • Market2Dish

    With the improvement of living standards, people's requirements for healthy diet are gradually increasing. People are paying more attention to the diversity, healthiness, balance and rationality of daily diets. Therefore, the demand for personalized health-aware food recommendation becomes more important and urgent. However, this task has its own challenges: 1) it’s non-trivial to learn a mapping function between the expected food and the huge ingredients in markets; 2) how to obtain user’s health profile needs study; 3) leveraging the existing knowledge and people’s health profile to recommend the personalized healthy food is a tough issue. To tackle these problems, we propose a health-aware food recommendation scheme, consisting of recipe retrieval, user health profiling, and health-aware food recommendation. In addition, we present this project as an online demonstration system. You can find the data, code and our demo here:

  • Healthcare System

    I lead a team to design and develop a community-based health system, named wenzher. To better serve health seekers, it automatically crawls and organizes multiple heterogeneous health-related sources. With state-of-the-art learning models, it intelligently mines the data insights and supports the following functions: deep text search, deep visual search, doctor recommendation, terminology annotation, symptom checker, disease report generation and other high-order analytics. It is publicly accessible via or an open version

  • Learning from Multiple Social Networks

    With the proliferation of social network services, more and more social users, such as individuals and organizations, are simultaneously involved in multiple social networks for various purposes. In fact, multiple social networks characterize the same social users from different perspectives, and their contexts are usually consistent or complementary rather than independent. Hence, as compared to using information from a single social network, appropriate aggregation of multiple social networks offers us a better way to comprehensively understand the given social users. My book on this project is available in Amazon. book

    We currently continue this research direction on learning from overlapping social networks and group profiling across multiple social networks.