Biography

I grew up in the east part of China and finished my undergraduate study there. At that time my major was about electric power transmission which is total different from AI. However, I was able to learn some knowledge about machine learning in some extracurricular activities. From then on I was fascinated by ML and AI. I earned my master's degree at the ECE department in University of Arizona, mainly studying artificial neural networks. Then I joined CS department for my PhD at UA, focusing on Natural Language Processing.

​​​​​Research Interests

My major interest is Question Answering (QA). Some QA tasks are really hard because they require the agent to acquire knowledge through natural language text, store the knowledge in memory and perform inference on top of the knowledge. None of these questions have been perfectly resolved so far. On the other hand, if an agent can fulfill the works above, then it is likely to be able to communicate with human in a natural way, which is a great progress in artificial intelligence. Therefore QA is a good test field of artificial agent and this is why it attracts me so much.

What long-term project do you want to work on?

The long-term goal is to come up with a neural architecture with external memory and reasoning module, which enables the agent to store and utilize knowledge more effectively. Moreover, it is preferable to let the agent learn from multiple types of inputs (e.g., video signals, audio signals, actions, etc.) except for text, just like human's learning from experience.

What do you enjoy most about your work?

The most fascinating part is to build an intelligent agent by writing each line of code. In addition, talking with the brilliant faculties and colleagues as well as attending academic presentations is really exciting.

What are your career goals?

The career goal is to build an agent which is similar to human.

​​​​Tell us something interesting about yourself!

During my spare time I watch movies and play videos games. But the most interesting thing is still research! :)