Towards an Approximative Ontology-Agnostic Approach for Logic Programs

Abstract

Distributional semantics focuses on the automatic construction of a semantic model based on the statistical distribution of co-located words in large-scale texts. Deductive reasoning is a fundamental component for semantic understanding. Despite the generality and expressivity of logical models, from an applied perspective, deductive reasoners are dependent on highly consistent conceptual models, which limits the application of reasoners to highly heterogeneous and open domain knowledge sources. Additionally, logical reasoners may present scalability issues. This work focuses on advancing the conceptual and formal work on the interaction between distributional semantics and logic, focusing on the introduction of a distributional deductive inference model for large-scale and heterogeneous knowledge bases. The proposed reasoning model targets the following features: (i) an approximative ontology-agnostic reasoning approach for logical knowledge bases, (ii) the inclusion of large volumes of distributional semantics commonsense knowledge into the inference process and (iii) the provision of a principled geometric representation of the inference process.