Jay Pujara, a research assistant professor of computer science at the USC Viterbi School of Engineering and the director of the Center on Knowledge Graphs at the USC Information Sciences Institute, recently received the 2023 Semantic Web Science Association Ten Year Award for his paper “Knowledge Graph Identification.” His paper won Best Student Paper in 2013 and is now again being honored with the SWSA Ten Year Award, which recognizes high-impact papers from the International Semantic Web Conference proceedings a decade earlier.
“It is a great honor to be considered as somebody who has made a very authoritative and long-lasting contribution to the community,” said Pujara, adding that Hui Miao, Lise Getoor and William Cohen co-authored the winning paper.
Said Adam Russell, director of ISI’s Artificial Intelligence division: “Jay’s work continues to lay important foundations for what we call ‘AI Next’ – building new approaches to AI we’ll need to tackle hard problems. And that’s going to take knowledge and wisdom. Jay has both.”
Knowledge graphs (KGs) connect entities like people, places, or things with other entities. In establishing a structured relationship between them, information organizes itself into a graph, enabling efficient data organization and retrieval. KGs are everywhere. If I looked up “Brazil” on a search engine like Google, results about the country’s location, proximity to me, and history would instantly pop up. Companies like Netflix and Amazon leverage KGs to make recommendations based on user habits and preferences.
Knowledge graph identification, the process of improving the quality of KGs by removing noisy information, inferring missing facts, and learning missing connections using a probabilistic model, results in higher quality KGs. Additionally, knowledge graph identification has real-world applications, especially in a future that will integrate more and more AI technologies. This is why Pujara’s paper remains relevant a decade after its publication.
“If you could build these clean and consistent knowledge graphs, it would help clients of these knowledge graphs get the information they need more quickly and with more confidence,” Pujara said.
Moreover, KGs are used in many ways beyond user-facing applications.
“These knowledge graphs are being used to make decisions like where we should invest resources to reduce food insecurity and poverty” he added.
Pujara continues to work with KGs and their applications as the director of the Center of Knowledge Graphs at ISI. Being one of the few universities in the country to have entire center devoted to the study of KGs, USC has uniquely facilitated Pujara’s and his colleagues’ research.
“The idea that there’s even a center of people who are excited about knowledge graphs and using them has really created this critical mass at USC around knowledge graphs,” said Pujara. “Centers and interest groups with a lot of motivated and engaged individuals are quite inspiring.”
Published on November 9th, 2023
Last updated on November 13th, 2023