Research & Development

Our ultimate R&D goals are to use user-generated content from various platforms and to employ the latest development from data analytics, machine learning, AI, and big data to benefit the society as a whole.

Data Mining for Finance:

For a long time, micro-blogging website Twitter has become a large community for finance investors to discuss and share information about stocks for better investment decisions. The social media, however, is also an ideal place in which many traders use “pump-and-dump” tactics to falsify information about the listed companies to illegally benefit themselves. We collected data from Twitter and analyze it using natural language processing to detect those activities.

Data Mining for Public Health and Social Science:

Thanks to the widespread of social media and mobile devices, every social event from anywhere is instantly reflected on the internet. The events range from a political story, a school bullying, to the outbreak of an emerging disease. We exploit the vast volume of publicly available data from the social media to provide the public health monitor in the population. The R&D can potentially lead to an Internet-based public health surveillance system with tremendous benefits such as a large population coverage, a larger demographic coverage, low cost, and low delay.

Evaluating marijuana-related tweets on Twitter " 420 Friendly": Revealing Marijuana Use via Craigslist Rental Ads
Machine Learning for Information Flow Optimization:

Reliable broadcasting data to multiple receivers over lossy wireless channels is challenging due to the heterogeneity of the wireless link conditions. Network coding (NC) has been shown to be a promising technique for improving network bandwidth efficiency by combining multiple lost data packets for retransmission. However, it is challenging to accurately determine which lost packets should be combined together due to disrupted feedback channels. We propose an adaptive data encoding scheme at the transmitter by joining network coding and machine learning for retransmission of lost packets. We have conducted extensive simulations to collaborate the efficiency of our proposed approach. The simulation results show that our machine learning algorithm can be trained efficiently and accurately. It achieves significant bandwidth gain compared with the existing schemes in different transmission terrains, power levels, and the distances between the transmitter and receivers.

Joint network coding and machine learning for error-prone wireless broadcast
Learn English by Games:

The team Saolasoft has successfully developed mobile application Learn English by Games during early 2016. Learn English by Games is an educational gaming app that enables the users to learn English via playing games with their friends with several creative features. One feature is that the content is generated and voted by the users. By using this crowd-source content, the users always find fresh and real life-related content in the game, rather than fixed one in all other English language learning applications. The other features involve users can send messages and exchange their English skills with their friends by a real- time notification system. Learn English by Games now has more than 80,000 registered users.