The Top Down Query Evaluation Engine for DataLog

Recursive query processing methods can be broadly categorized as either bottom-up methods, or top-down methods. Bottom-up methods answer a query by applying all rules of a program to ground tuples, deriving tuples that satisfy rule bodies into predicates in rule heads. The minimal model for the given program and ground tuples is explicitly materialized as a new database instance; the answer to the query is then obtained through a simple select/project/join operations over the materialized database instance. In contrast, top-down methods answer a query by pushing selection criteria (i.e. constants) from the query down into rules that may answer the query (i.e. rules deriving into predicates being queried), creating more (sub)queries from the atoms of these rules’ bodies; the subqueries are in turn answered in a sim- ilar, top-down fashion. In this project, we implement a representative method for the top-down evaluation : query/subquery (QSQ).

Hana Sebia
Hana Sebia
PhD Candidate in Machine Learning

My research interests include healthcare data analytics, deep learning, tensor factorization and predictive analysis.