My research interests include Machine Translation, Chinese Language Processing, Natural Language Understanding, Statistical Language Modeling and Machine Learning. I have worked on statistical modeling and machine learning for natural language processing since 1986 (mainly adopting principled approaches to solving NLP problems, in contrast to the prevailing techniques of simply adding a lot of features under the Maximum Entropy framework).

My past experience mainly lies in building a statistical semantic machine translation system. I had proposed and built a statistics-oriented framework for machine translation, which automatically obtains associated parameters from the given pairs of Source and Target Semantic-Normal-Forms via semi-supervised learning. Besides, I had proposed a joint model to integrate Translation Memory into a phrase-based machine translation system during decoding, which has achieved significant progress.

I have also worked on Chinese word segmentation with new generative models, and have obtained the best performance in the literature on SIGHAN close/open-test (which is classified according to whether additional resources such as dictionaries other than the training-set data could be adopted).

I will start to build an Intelligent Q&A System which is able to give the desired answer that cannot be directly fetched from the given text. In other words, we would like to attack the problem that inference is required in getting the answer (or the result must be aggregated from the information distributed in various locations). In addition to the answer, the associated inference chain (and the corresponding confidence level) will also be given. This project is expected to start from reading elementary school text first, gradually shift to high school, and then to real domain-oriented applications. The progress of this project will be evaluated by the comprehensive test associated with the given text.