The Zoom link for this lecture will be posted here on the day of the event, as well as sent to the CS 201 email list.

“Fraternal Twins: The Long History of Tables and Tensors”

Going all the way back to 1870, researchers have recognized the uncanny similarity between relations and tensors. This similarity has resulted in an intermittent stream of work that attempts to bridge the two fields in both systems and theory. Database theoreticians have proposed languages like MatLang then studied their expressiveness as compared to relational algebra. HPC researchers have attempted to handle database operations via PyTorch and Onyx.  Database systems engineers have desperately attempted to capture ML workloads by 1) incorporating array operations into databases or 2) building database-style optimizers for linear and tensor algebra. Meanwhile, algorithm theoreticians have begun to borrow database theory techniques to analyze the complexity of sparse tensor algebra kernels. In this talk, I will attempt to provide a history of the interaction between these fields, establish a basic correspondence between their data models, and discuss the key systems techniques that each field brings to the table.

Related Work (More for skimming than a deep read):

Galley: Modern Query Optimization for Sparse Tensor Algebra
Query Processing on Tensor Computation Runtimes
Expressive Power of Linear Algebra Query Languages

Kyle Deeds is an assistant professor of computer science at Boston University. His work focuses on the analysis and optimization of query languages like relational algebra and sparse tensor algebra. He is a lead developer on the Finch sparse array framework. In addition, he is excited about open data initiatives and co-leads the govscape.net project. He has been published in a variety of data management conferences including SIGMOD, VLDB, PODS, and ICDT, where he won the best paper award in 2025. He was also a runner up for the ACM SIGMOD Jim Gray Dissertation Award in 2026.