LLMs can dynamically query Knowledge Graphs to generate contextual, domain-aware explanations of ML model predictions—making AI decisions more transparent and trustworthy in specialized industries like manufacturing.
This paper combines Knowledge Graphs and Large Language Models to explain machine learning predictions in manufacturing. The system stores domain knowledge and ML results in a structured graph, then uses an LLM to convert relevant information into clear, user-friendly explanations. Testing shows the approach works well for both standard and complex questions in real manufacturing settings.