*********************************
There is now a CONTENT FREEZE for Mercury while we switch to a new platform. It began on Friday, March 10 at 6pm and will end on Wednesday, March 15 at noon. No new content can be created during this time, but all material in the system as of the beginning of the freeze will be migrated to the new platform, including users and groups. Functionally the new site is identical to the old one. webteam@gatech.edu
*********************************
Title: Symbolic Reasoning for Query Verification and Optimization
Qi Zhou
School of Computer Science
Georgia Institute of Technology
Date: Tuesday, November 24, 2020
Time: 12:00 pm - 3:00 pm (ET)
Location: remote, via BlueJeans at https://bluejeans.com/259615650
Committee:
Dr. William Harris (advisor) - Galois Inc.
Dr. Joy Arulraj (co-advisor) - School of Computer Science, Georgia Institute of Technology
Dr. Shamkant B.Navathe - School of Computer Science, Georgia Institute of Technology
Dr. Alex Orso - School of Computer Science, Georgia Institute of Technology
Dr. John Regehr - School of Computing, University of Utah
Abstract:
Standard Query Language (SQL) is the most widely used language for interacting with many database management systems (DBMS). Thus, the problems of optimizing and verifying SQL queries are the most studied problems in the DBMS community. Traditional techniques for optimizing and verifying SQL queries are based on syntax-driven approaches, which suffer many limitations in terms of effectiveness and efficiency.
In this dissertation, I investigate two important problems in query verification and optimization to demonstrate the limitations of syntax-driven techniques: (1) proving query equivalence under set and bag semantics; (2) optimizing queries with learned predicates. I propose to use symbolic reasoning to address the limitations of syntax-driven approaches in these two problems. I first present two techniques for proving query equivalence under set and bag semantics based on symbolic representation. Both approaches are significantly more efficient and effective than the previous state-of-the-art syntax-driven techniques. I then present a novel algorithm that combines symbolic reasoning with machine learning to synthesize new predicates for optimizing queries. This algorithm enables the query optimizer to leverage more optimization rules that it cannot previously apply. This technique significantly speeds up the execution of queries with complex predicates. In conclusion, this thesis proved that using symbolic reasoning can significantly improve the efficiency and effectiveness of techniques for query equivalence verification and query optimization.
