Stratified Knowledge Bases as Interpretable Probabilistic Models (Extended Abstract)

TL;DR

This paper proposes using stratified logical theories to create more interpretable probabilistic models, enabling domain experts to directly modify models for improved accuracy and clarity.

Contribution

It introduces a novel approach of representing probabilistic models with stratified logical theories for enhanced interpretability and expert intervention.

Findings

Stratified logical theories improve model interpretability.

Domain experts can directly modify models by editing logical formulas.

The approach offers a transparent alternative to Markov logic networks.

Abstract

In this paper, we advocate the use of stratified logical theories for representing probabilistic models. We argue that such encodings can be more interpretable than those obtained in existing frameworks such as Markov logic networks. Among others, this allows for the use of domain experts to improve learned models by directly removing, adding, or modifying logical formulas.