COMPASS Talks
The Computing Platforms Seminar Series (COMPASS) is focused on talks by industry and academia around the general topic of computing platforms.
Filters that Expand to Infinity and Beyond
Abstract:
Filter data structures are widely used to answer approximate set-membership queries in various areas of computer science. A filter needs to be allocated with a given capacity in advance to provide a guarantee over the false positive rate and performance. However, In many applications, the data size is not known in advance, requiring filters to expand dynamically. We show that existing methods for expanding filters do not scale in terms of performance or the false positive rate. This talk will cover InfiniFilter (SIGMOD 2023) and Aleph Filter (VLDB 2024), two new expandable filters that address these challenges.
Short Bio:
Niv Dayan is an assistant professor at the University of Toronto. He is interested in designing and analyzing data structures for database management and storage systems. Before joining the University of Toronto, Niv was a research scientist at Pliops, a startup working on storage engine accelerators. He was a postdoc at Harvard and Copenhagen University, and his PhD is from the IT University of Copenhagen.