A Rule-Based Approach to Specifying Preferences over Conflicting Facts and Querying Inconsistent Knowledge Bases

TL;DR

This paper tackles querying inconsistent knowledge bases by introducing a declarative, rule-based framework to specify a priority relation over conflicting facts. It analyzes the acyclicity properties of the induced preference relation and offers pragmatic cycle-resolution techniques to obtain an acyclic ordering, enabling prioritized repairs under AR, IAR, and Brave semantics. An end-to-end ASP-based implementation is presented, including modules for data/metadata encoding, conflict detection, priority computation, and semantics-driven query answering, with a public repository for reproducibility. Experiments on DL-Lite benchmarks demonstrate both the feasibility and the scalability challenges, showing favorable performance for certain cycle-resolution strategies and highlighting trade-offs between non-binary conflicts and existing SAT-based approaches.

Abstract

Repair-based semantics have been extensively studied as a means of obtaining meaningful answers to queries posed over inconsistent knowledge bases (KBs). While several works have considered how to exploit a priority relation between facts to select optimal repairs, the question of how to specify such preferences remains largely unaddressed. This motivates us to introduce a declarative rule-based framework for specifying and computing a priority relation between conflicting facts. As the expressed preferences may contain undesirable cycles, we consider the problem of determining when a set of preference rules always yields an acyclic relation, and we also explore a pragmatic approach that extracts an acyclic relation by applying various cycle removal techniques. Towards an end-to-end system for querying inconsistent KBs, we present a preliminary implementation and experimental evaluation of the framework, which employs answer set programming to evaluate the preference rules, apply the desired cycle resolution techniques to obtain a priority relation, and answer queries under prioritized-repair semantics.

Figures (3)

• Figure 1: Time (in sec.) to compute $\succ^x$ from the pre-computed conflicts for $\mathsf{u1cY}$ given as a program $\Pi_{\mathit{conf}}$ and $\Pi_\mathcal{D}\cup\Pi_\mathcal{F}\cup\Pi_P\cup\Pi_{\succ^x}$ in scenarios (a) and (b) w.r.t. the number (in thousands) of conflicts. Empty bars for $\mathsf{u1c50}$ (81K conflicts) mean t.o. or oom. The lower part of each bar (light grey) shows the time to ground the ASP program while the upper part is the time to solve it.
• Figure 2: Additional denial constraint: members of department 14 of university 0 cannot take three distinct graduate courses on subject 11.
• Figure 3: Time (in sec.) to compute $\succ^x$ from the pre-computed conflicts given as a program $\Pi_{\mathit{conf}}$ and $\Pi_\mathcal{D}\cup\Pi_\mathcal{F}\cup\Pi_P\cup\Pi_{\succ^x}$ in scenarios (a), (b), (c) and (d) w.r.t. the number (in thousands) of conflicts for $\mathsf{u1cY}$ (75-78K facts). The lower part of the bars (light grey) is the time to ground the program while the upper part is the time to solve it. Empty bars mean a time-out or oom.

Theorems & Definitions (48)

• Definition 1
• Definition 2
• Example 1
• Theorem 1
• Theorem 2
• Example 2
• Remark 1
• Definition 3
• Example 3
• Definition 4