Year: 2025
Type: conference-paper
Abstract: We explore the landscape of modern data structures that employ two powerful scaling mechanisms simultaneously: (a) concurrency to harness the power of today's multicore processors, and (b) learned queries -the operations employing machine learning inference by models trained on the underlying data.We focus on advancements, challenges, and potential future directions in this field.We delve into the motivation for incorporating learned queries in concurrent data structures.We discuss the key techniques, applications, and open research questions.On the theoretical side, we discuss the mathematical foundations for building the learned data structures in addition to the correctness of queries in a concurrent setting.On the practical side, we present an overview of techniques used for managing concurrency with learned queries that need training the models with concurrent real-time updates.
Authors: Bapi Chatterjee, Sathya Peri
Language: English
Subfield: Information Systems
Field: Computer Science
Domain: Physical Sciences
Open Access status: hybrid
OpenAlex ID: W4405992896