The Open Virtual Assistant Initiative

Zoom Link

Passcode:  421546

Abstract: This talk describes our initiative at Stanford to create an open-source alternative to commercial virtual assistants, with the goals of protecting privacy and keeping the voice web open.  Virtual assistants today rely mainly on training neural networks with manually annotated natural language utterances, which is prohibitively expensive and ineffective.  This talk presents a new methodology for developing conversational agents: training with mostly data synthesized with the help of pertained language models. This approach has led to natural language semantic parsers that can understand long-tail questions on structured data better than commercial assistants and can be ported to other languages in a day.

Bio: Dr. Monica Lam has been a Professor of Computer Science at Stanford University since 1988, and is the Faculty Director of the Stanford Open Virtual Assistant Laboratory.  She leads the Genie open virtual assistant project, aimed to advance and democratize voice assistant technology, keep the voice web open, and protect the privacy of consumers.

Prof. Lam is a member of the National Academy of Engineering and an ACM Fellow.  She has won numerous best paper awards, and has published over 150 papers on many topics: natural language processing, machine learning, compilers, computer architecture, operating systems, high-performance computing, and HCI.  She is an author of the “Dragon Book”, the definitive text on compiler technology.  She received a B.Sc. from University of British Columbia (1980) and a Ph.D. from Carnegie Mellon University (1987).