Research Fellow  |  Chen, Sheng-Wei (Kuan-Ta)  
Personal (New window)
lab (New window)
Research Descriptions

My research areas span over big data analytics, deep learning, multimedia systems and social computing. Though these fields the two fields look unrelated, they are now collectively part of modern life and affecting what people do and think in countless ways. For example, demonstrations and any social events can now be watched and streamed from any perspective (as long as there is a smartphone) over the Internet and the opinions can be disseminated and aggregated with an unprecedented speed on social information systems. The advances in multimedia and social computer systems have drastically changed how we sense, interpret, and respond to other people and the society. 

Big Data Analytics and Deep Learning

Data science is an interdisciplinary field about theories, techniques, processes, and systems to extract knowledge or insights from data in various forms, either structured or unstructured. It is highly related to fundamental fields such as statistics and applied fields such as data mining, pattern recognition, and knowledge discovery in databases, as well as big data engineering/analytics. It seems that data science simply old wine in a new bottle, whereas the fact that a number of new techniques and tools have been invented to resolve issues in analytics and engineering provides a contradiction.

Social Computing and Computational Social Science

The strongest incentive that drive me to get into the social computing field is the concern of social welfare. I''d say my research on social computing are around poverty and happiness, as poverty is one of the biggest social problem in the world (and unfortunately becoming worse over time), while people who are not poor may still be unhappy due to various social factors, such as anxiety disorders, relative deprivation, social disparity, alienation, and so on.  As we can see from the human history, the technology advancement and economic growth will not resolve the huge, deep, unbearable problem of humans --- how to live our lives happily?  I am working with sociologists and psychologists with a hope to address the social well-being problem from an interdisciplinary perspective.


Multimedia Systems

My focus on multimedia systems has been on real-time interactive networked multimedia systems, particularly the audio/video conferencing systems and the network gaming systems, as the general public uses the Internet for two primary reasons: interpersonal interaction and entertainment.  

It has been more than 20 years since audio/video communication over the Internet has been made possible. Still, remote communication remains much less natural and efficient than face-to-face communication. People still fly to office branches, customers'' companies, and conferences to attend meetings whenever possible for an obvious reason´╝Źthe current systems only allow us to do the most simple visual and verbal communication, whereas more subtle social signals, such as proximity, direction of gesture, eye contact are not communicable via the current systems.

My motivations for developing an affordable immersive remote communication system are multi-fold: 1) We believe that such a system would largely reduce the demand of business flights, which in turn reduces travel costs and fatigue, and also helps to protect the global environment; 2) in some extreme circumstances, such as during natural disasters, travel would not be possible, but rich remote communication is definitely required; and 3) we believe that richer and more immersive remote communication would help people to understand and collaborate with each other in more depth (imagining that a politician can have eye contacts with ten thousands of audience at the same time when giving a speech over the Internet). While most problems in this world might be due to misunderstanding, we hope that better remote communication tools would help people across the globe understand each other and the world better and create a world with less conflicts, arguments, and wars.

I also have a keen interest in computer gaming since childhood, and my research on network gaming dates back to 2004 with the goal of providing a scalable, efficient, and secure gaming infrastructure for all kinds of network games. Network gaming is now one of the most profitable businesses on the Internet, and its market share has surpassed those of the movie and music industries. To ensure that this line of research is solid and practical, I have been collaborating with a number of game companies that provide their domain knowledge, real-life operation logs, and funding for research projects. My network game study covers the system design issues, including traffic analysis, transport protocols, and how to use cloud computing to better deliver game content; we also worked on security issues in network gaming, such as game bot detection and account hijacking detection.