[Paper Review] A Hybrid Approach to Query Answering under Expressive Datalog+/-

Mostafa Milani, Andrea Calı̀+1 more|arXiv (Cornell University)|Apr 22, 2016

Semantic Web and Ontologies14 references4 citations

TL;DR

This paper presents two novel deterministic algorithms for conjunctive query answering in weakly-sticky (WS) Datalog±, a highly expressive ontology language. The first uses a query-driven, resumption-based chase to generate a query-specific database expansion; the second employs a hybrid transformation into a sticky program followed by standard rewriting. Both achieve optimal data complexity (ptime), enabling efficient, practical query answering in real-world OBDA scenarios.

ABSTRACT

Datalog+/- is a family of ontology languages that combine good computational properties with high expressive power. Datalog+/- languages are provably able to capture the most relevant Semantic Web languages. In this paper we consider the class of weakly-sticky (WS) Datalog+/- programs, which allow for certain useful forms of joins in rule bodies as well as extending the well-known class of weakly-acyclic TGDs. So far, only non-deterministic algorithms were known for answering queries on WS Datalog+/- programs. We present novel deterministic query answering algorithms under WS Datalog+/-. In particular, we propose: (1) a bottom-up grounding algorithm based on a query-driven chase, and (2) a hybrid approach based on transforming a WS program into a so-called sticky one, for which query rewriting techniques are known. We discuss how our algorithms can be optimized and effectively applied for query answering in real-world scenarios.

Motivation & Objective

• To address the lack of deterministic algorithms for conjunctive query answering in weakly-sticky (WS) Datalog±, a highly expressive yet decidable ontology language.
• To overcome the reliance on nondeterministic chase procedures in prior work, which hinders practical deployment.
• To develop efficient, scalable query answering techniques that maintain decidability while supporting complex joins and existential rules.
• To enable practical deployment in real-world ontology-based data access (OBDA) systems through optimized, deterministic algorithms.
• To achieve optimal data complexity (ptime) for query answering under WS Datalog±, matching the theoretical lower bound.

Proposed method

• Introduces a query-driven, resumption-based bottom-up chase algorithm that expands the database in steps corresponding to the number of existentially quantified variables in the query.
• Applies a two-phase transformation: first, Skolemization and annotation to eliminate finite-rank existential variables, then partial grounding on zero-ranked variables to produce a sticky program.
• Employs standard rewriting techniques for sticky Datalog± to transform the original query into a union of conjunctive queries (UCQs).
• Evaluates the rewritten UCQs directly over the original database instance, enabling efficient SQL-style execution.
• Uses dependency graph analysis to optimize grounding by avoiding unnecessary rule expansions when certain constants do not yield new atoms.
• Combines off-line pre-computation of expansions (for up to m existential variables) with on-demand resumption to support incremental query processing.

Experimental results

Research questions

• RQ1 Can deterministic, efficient query answering be achieved for the expressive WS Datalog± language, which extends both sticky and weakly-acyclic Datalog±?
• RQ2 How can the chase procedure be modified to ensure termination and determinism while remaining query-driven and efficient?
• RQ3 Can a hybrid transformation into a sticky program preserve query answers while enabling use of existing, efficient rewriting techniques?
• RQ4 What is the data complexity of conjunctive query answering under WS Datalog±, and can it be achieved in ptime?
• RQ5 How can grounding be optimized to avoid generating redundant atoms during transformation?

Key findings

• The proposed query-driven chase algorithm achieves optimal data complexity (ptime) for conjunctive query answering under WS Datalog±, with the expansion process being deterministic and query-specific.
• The hybrid approach transforms a WS program into a sticky program via Skolemization and partial grounding, enabling the use of existing efficient rewriting techniques for sticky Datalog±.
• Both algorithms achieve the optimal ptime lower bound in data complexity, confirming their theoretical efficiency and practical viability.
• The first algorithm supports incremental query processing by pre-computing expansions for queries with up to m existential variables, allowing reuse and on-demand extension.
• The second algorithm enables efficient evaluation via standard SQL-like processing of rewritten UCQs over the original database, facilitating integration with RDBMS.
• Optimization via dependency graph analysis can reduce the number of generated CQs in the rewriting phase by avoiding unnecessary grounding steps.