More about HKUST
CONJUNCTIVE QUERY PROCESSING WITH COMPARISONS AND UPDATES
PhD Thesis Proposal Defence Title: "CONJUNCTIVE QUERY PROCESSING WITH COMPARISONS AND UPDATES" by Mr. Qichen WANG Abstract: Conjunctive queries form the backbone of SQL queries, and have been extensively studied in the literature. In this proposal, we consider two important extensions of a basic conjunctive query. First, we study how to evaluate conjunctive queries with predicates in the form of comparisons that span multiple relations. Such predicates have regained interest, due to their relevance in OLAP systems, spatiotemporal databases, and machine learning over relational data. We design a new algorithm for evaluating conjunctive queries with comparisons and identify an acyclic condition under which linear query processing time can be achieved. We also extend our acyclic query algorithm to more general queries, resulting in polynomial-factor improvements over previous best results for many cyclic queries. Secondly, we study the problem of incrementally maintaining the results of a conjunctive query under updates. Prior work has shown that this problem is inherently hard in the worst case. To provide an instance-specific analysis of the update cost, we introduce a new measure of complexity of the update sequence, which we call the enclosureness, which is a small constant for many realistic update sequences, such as first-in-first-out (FIFO) sequence. We present an algorithm to maintain the query results of any acyclic foreign-key join query in time linear in the enclosureness. We are currently working on extending this algorithm to non-key joins. All newly proposed algorithms are efficient not only in theory but also in practice. We have implemented them on Spark and Flink, and our experimental results demonstrate order-of-magnitude speedups against the previous state-of-the-art results. Date: Friday, 18 February 2022 Time: 2:00pm - 4:00pm Zoom Meeting: https://hkust.zoom.us/j/95281613710?pwd=ZUJBcENNUUh3SnU4akpRRDUxU24rQT09 Committee Members: Prof. Ke Yi (Supervisor) Prof. Lei Chen (Chairperson) Prof. Qiong Luo Prof. Xiaofang Zhou **** ALL are Welcome ****