COSIC Seminar "Privacy-Preserving Graph Analysis: The Journey So Far and..." (Bhavish Raj Gopal) скачать в хорошем качестве

COSIC Seminar "Privacy-Preserving Graph Analysis: The Journey So Far and..." (Bhavish Raj Gopal)

COSIC Seminar - Privacy-Preserving Graph Analysis: The Journey So Far and the Way Ahead - Bhavish Raj Gopal (Indian Institute of Science) Abstract: Graphs are powerful tools used to model relationships between entities, but in many real-world scenarios—such as social networks, supply chains, and financial systems—the graph data is distributed across multiple organizations and contains sensitive information. Analyzing such graphs jointly can yield valuable insights, yet directly sharing data is often prohibited due to privacy, regulatory, or competitive concerns. In this talk, we focus on how secure multiparty computation (MPC) can be used to address this challenge. MPC is a cryptographic technique that allows 𝑛 mutually distrusting parties to jointly compute a function over their private inputs while revealing nothing beyond the output. In the considered scenario, the private input is the fragment of graph that is distributively held by each data owner, while the function of interest is a graph algorithm such as PageRank, BFS, or connected components. We will introduce two MPC-based frameworks—Graphiti and emGraph, each designed to support different practical deployment settings. The talk will include key technical ideas underlying these frameworks, practical performance benchmarks, and potential real-world applications in areas such as fraud detection and contact tracing. Through this, we demonstrate that privacy-preserving graph analysis can be made scalable and practical. Bio: Bhavish Raj Gopal is a Ph.D. candidate in the Department of Computer Science and Automation at the Indian Institute of Science (IISc), where he is advised by Prof. Arpita Patra. He holds an integrated BS-MS in Mathematics from the Indian Institute of Science Education and Research (IISER) Mohali and is a recipient of the Prime Minister’s Research Fellowship for his doctoral studies. Bhavish’s research focuses on the applied aspects of secure multiparty computation (MPC), with an emphasis on designing privacy-preserving solutions for real-world applications, such as machine learning, private heavy hitters, pattern matching, and graph algorithms, to name a few. His research has been published in top-tier venues including CCS, TheWebConf, PoPETs, and IEEE TDSC.