Aiming at the problems of multi-source heterogeneous data and inconsistent product quality consistency in the outsourcing processing of papermaking machinery wallboard under the cloud manufacturing paradigm, this paper proposed a method of building the outsourcing model of wallboard based on knowledge graph. The method analyzed the processing information such as process parameters and equipment, as well as their inherent relationships. An ontology modeling strategy was employed, combining the seven-step method and skeleton method.
An improved BiLSTM-CRF extraction algorithm was introduced, along with a BERT-based cosine similarity algorithm for entity semantic coreference resolution. Neo4j graph database was used for knowledge storage and graph construction. The results showed that the proposed extraction model outperformed the comparison models in terms of precision, recall, and F1-score on the self-created dataset, significantly improving the efficiency of information flow and decision support during the outsourcing processing process.