Chinese Entity Recognition for Cloud Manufacturing Knowledge Graph Construction

To address the lack of specific knowledge base support for manufacturing service retrieval and resource allocation operations in cloud manufacturing environments, a two-step approach has been proposed to construct a comprehensive knowledge base. Firstly, ontology construction methods are employed to determine entity categories, knowledge classification criteria, and annotation strategies. Secondly, manufacturing resource and manufacturing task corpus datasets are meticulously constructed. Furthermore, a novel Integration of Ontological Features (IOF) approach, combining rule-based and conditional random field-based techniques (IOF-Bi-LSTM-CRF), is introduced for manufacturing task entity recognition. This approach leverages a bidirectional long short-term memory (BiLSTM) network to extract vector features, and utilizes conditional random field (CRF) to derive optimal results and assign labels to entity categories. Experimental results demonstrate that the proposed method achieves a recognition accuracy of over 85% for the manufacturing task entity category. It successfully identifies key entities within Chinese text of manufacturing tasks in the cloud manufacturing environment, thereby establishing a solid foundation for the automated construction of a cloud manufacturing knowledge graph.