Braid: Weaving Symbolic and Neural Knowledge into Coherent Logical Explanations
This work addresses the problem of brittle inference and knowledge acquisition in symbolic reasoning for researchers and practitioners seeking more robust and explainable AI systems.
This paper introduces Braid, a novel logical reasoner that integrates symbolic and neural knowledge to overcome the limitations of traditional symbolic reasoning, such as brittle inference and the knowledge acquisition problem. Braid supports probabilistic rules, custom unification functions, and dynamic rule generation, achieving close to state-of-the-art results on the ROC Story Cloze test while providing frame-based explanations.
Traditional symbolic reasoning engines, while attractive for their precision and explicability, have a few major drawbacks: the use of brittle inference procedures that rely on exact matching (unification) of logical terms, an inability to deal with uncertainty, and the need for a precompiled rule-base of knowledge (the "knowledge acquisition" problem). To address these issues, we devise a novel logical reasoner called Braid, that supports probabilistic rules, and uses the notion of custom unification functions and dynamic rule generation to overcome the brittle matching and knowledge-gap problem prevalent in traditional reasoners. In this paper, we describe the reasoning algorithms used in Braid, and their implementation in a distributed task-based framework that builds proof/explanation graphs for an input query. We use a simple QA example from a children's story to motivate Braid's design and explain how the various components work together to produce a coherent logical explanation. Finally, we evaluate Braid on the ROC Story Cloze test and achieve close to state-of-the-art results while providing frame-based explanations.