Forough Arabshahi

Carnegie Mellon University

PhD Candidate

Forough is currently a postdoctoral associate supervised by Tom Mitchell in the Machine Learning Department of Carnegie Mellon University. In her PhD, she was supervised by Anima Anandkumar and Sameer Singh where she worked on developing probabilistic and deep learning models for learning latent hierarchical structures. Prior to that, she received her M.Sc. from Amirkabir University of Technology, Tehran, Iran, and her B.Sc. from Shiraz University, Shiraz, Iran.

Research Abstract:

Ever since the rise of artificial intelligence (AI) in the twentieth century, researchers have dreamed about creating a computer agent whose intelligence is comparable to that of a human. The field of AI and machine learning have recently gone through a remarkable revolution. Together with the recent advances in natural language processing, we now have intelligent assistants such as Alexa, Siri and Cortana to help us with our daily activities. Even though these assistants are good at performing a set of pre-defined tasks within limited domains, they are still far from truly intelligent agents and still lack commonsense reasoning abilities. The goal of my research is to develop a conversational AI assistant that is capable of performing commonsense reasoning. Conducting research in this area requires a diverse set of skills in probabilistic learning, deep learning and natural language processing.