Weighted Conditional EL{^}bot Knowledge Bases with Integer Weights: an ASP Approach

TL;DR

This paper introduces an ASP-based approach for reasoning with weighted conditional EL^bot knowledge bases using integer weights, enhancing the logical semantics of multilayer perceptrons.

Contribution

It presents a novel encoding of concept-wise multipreference entailment for weighted KBs with integer weights using ASP and asprin.

Findings

Effective encoding of weighted KBs with ASP

Enhanced reasoning capabilities for concept-wise multipreference

Potential applications in neural-symbolic integration

Abstract

Weighted knowledge bases for description logics with typicality have been recently considered under a "concept-wise" multipreference semantics (in both the two-valued and fuzzy case), as the basis of a logical semantics of Multilayer Perceptrons. In this paper we consider weighted conditional EL^bot knowledge bases in the two-valued case, and exploit ASP and asprin for encoding concept-wise multipreference entailment for weighted KBs with integer weights.