OmEGa(Ω): Ontology-based information extraction framework for constructing task-centric knowledge graph from manufacturing documents with large language model

Midan Shim, Hyojun Choi, Heeyeon Koo, Kaehyun Um, Kyong-Ho Lee, Sanghyun Lee

Abstract: Highlights•Task-centric ontology is designed to describe fundamental manufacturing tasks.•Ontology-based Information Extraction Framework constructs task-centric KG.•Data augmentation alleviates contextual challenges of domain-specific document.•Spatial embedding relieves structure challenges of multimodal manufacturing document.•Experiments on both proprietary and open-source datasets show strong performance.