I am a Research Scientist at Facebook AI. Currently I am with the Language and Translation Technology team working on cross-lingual semantic representation, neural machine translation, large scale language data quality issues, and general NLP research.
Previously I was an Applied Scientist at Microsoft AI & Research with the CNTK team working on research and engineering for Microsoft's Deep Learning Toolkits CNTK. Before I was a project scientist at the Robotics Institute at the School of Computer Science at Carnegie Mellon University working on hybrid symbolic and probabilistic reasoning for AIs and human-AI interaction.
I am specialized in a broad range of AI technology across deep learning, multi-agent systems, AI planning, trust, and dialogues. I am in the program committees of a few prestigious AI conferences --- AAAI, EJCAI, ECAI, AAMAS, PRIMA, COMMA (Computational Model of Arguments) and other international conferences on various AI topics, such as trust and argumentation.
Currently I am focusing on Deep Learning and neural representation of (cross-lingual) language understanding and generation --- an area in which we might have a good chance to break through the latest deep learning advancement of perception AI into cognitive AI. However, my long term goal is towards the integration of human knowledge (e.g. free text wiki pages and structural knowledge bases) and data-driven machine learning (.e.g. probabilistic graphical models and deep learning). Along this line, my past work spans across probabilsitic graphical models, probabilistic symbolic reasoning (e.g. probabilistic argumentation-based reasoning), probabilistic multi-agent planning, and multi-agent dialogue planning, human-robot interaction, trust modeling in social networks, agent-based simulation and so on. In general, I am interested in AI techniques that can support the establishment of social intelligence for a system of agents (e.g. machine agents, human, or hybrid human-machine agents).