Stratified Knowledge Bases as Interpretable Probabilistic Models (Extended Abstract)
This addresses the need for more interpretable probabilistic models in AI, particularly for applications involving domain experts, but it appears incremental as it builds on existing frameworks like Markov logic networks.
The paper tackles the problem of interpretability in probabilistic models by proposing stratified logical theories as a representation, which allows domain experts to directly modify logical formulas to improve learned models.
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.