+ Director: Khoat Than
+ Funding agency: Air Force Office of Scientific Research, USA
+ Project title: Inferring the latent structures in big heterogeneous data
+ Main areas: Machine learning, Big data
Abstract: The ultimate goal in this project is to develop a class of inference algorithms that enable us to explore and discover hidden structures (semantics) from massive text collections, and that enable us to do accurate predictions in practical applications. Particularly, those algorithms should be able to easily handle text streams or textual collections of size millions. Such algorithms will have significant impacts in many areas as they enable us to understand/exploit a large amount of heterogeneous and dynamic texts. To this end, we base our methodologies on some key directions in Machine Learning including topic modeling, online learning, and stochastic inference. Our inference algorithms will be exploited to develop efficient methods for question answering, recommendation systems, social network analysis, and information retrieval.