Industry-Sponsored Research Week

Google doles out $1M in research funding for AI recommender systems to California researchers

By David Schwartz
Published: November 17th, 2020

Researcher from three California universities — UC Santa Cruz, UC Santa Barbara, and the University of Southern California — will receive grants from Google totaling $1 million over three years to support their research into online recommender dialogue systems.

Recommender systems are among the most familiar practical applications of artificial intelligence (AI), suggesting music and movies people might like and products they might want to buy based on their past purchasing or search behavior. Recommender dialogues, also known as conversational recommender systems, go a step further by engaging users in interactive dialogue to elicit more details about the their interests. But current dialogue agents have limited capabilities.

Lise Getoor, a computer scientist at UC Santa Cruz, and her collaborators at the other California campuses bring a mix of backgrounds and expertise to the project. William Wang, director of UCSB’s Natural Language Processing group and Center for Responsible Machine Learning, has broad interests in machine learning approaches to data science, including natural language processing and knowledge representation. USC’s Jay Pujara is a research assistant professor of computer science at USC whose AI research focuses on user utility.

“We form a strong tripod to support this project in terms of our backgrounds,” Getoor said. “We will also be working with leading experts at Google in machine learning, natural language processing, and knowledge-based reasoning.”

The team’s research will rely on Probabilistic Soft Logic (PSL), an open-source toolkit developed by Getoor’s group at UC Santa Cruz that has been applied to a wide range of machine learning problems. PSL is a highly scalable probabilistic programming framework that is able to reason both statistically, using similarities, and logically, using soft rules.

“PSL is different from deep learning because of the way you can express soft tendencies, such as rules that apply most of the time but not all of the time, so it’s very intuitive and its results are easier to interpret,” Getoor explained.

The researchers plan to extend PSL in key ways to integrate it with neural network-based deep learning approaches, which have gained popularity but have some significant limitations. “The core of the project is figuring out how to integrate these approaches. What that will enable is a more informed dialogue that takes into account more of the user’s context,” Getoor said. Wang added that the researchers hope to improve the reasoning capabilities of dialogue agents, so that the system can better understand a user’s request, using context and other information to generate a more appropriate response.

“For example, if I say I just had Italian for lunch, can you recommend a restaurant for dinner, a lot of dialogue systems don’t understand semantic constraints, so it might recommend a pizza place because of the implicit connection between pizza and Italian food,” Wang said. “That’s an issue that is often frustrating for users.”

The team will attempt to design better algorithms and models for understanding the roles of context, knowledge, and uncertainty in dialogues. In particular, the team will investigate natural language understanding, knowledge discovery and reasoning, and natural language generation in task-oriented dialogues – all key areas of to Google.

Getoor previously received a Google Faculty Research Award in 2019, and the new project builds on the work supported by that earlier grant.

Source: UC Santa Cruz Newscenter

Posted under: University-Industry Engagement Week

Twitter Facebook Linkedin Pinterest Email

No Comments so far ↓

There are no comments yet...Kick things off by filling out the form below.

Leave a Comment