Abstract: Efficient reuse of process knowledge is the key to alleviate the high dependence of remanufacturing process planning on manual labor. However, due to the multi-source heterogeneity of process entities and the inefficiency of knowledge retrieval, the reuse of remanufacturing process knowledge still faces significant challenges. To address these issues, this paper proposes an efficient reuse method for remanufacturing process knowledge based on multi-source heterogeneous entity alignment and key node matching. Firstly, based on the BERT-BiLSTM-CRF model and rule-based knowledge extraction, a hybrid entity alignment method integrating attribute similarity and embedding similarity is proposed to solve the problems of inconsistent triples extracted from different sources and knowledge redundancy. Subsequently, a remanufacturing process planning method based on similarity matching of failure key nodes is proposed to achieve efficient reuse of knowledge. Finally, the feasibility and superiority of this method are verified by a crankshaft remanufacturing case. Results show that the constructed knowledge graph in this paper realizes the structured storage and unified management of remanufacturing case data, also, the proposed hybrid algorithm significantly outperforms conventional attribute similarity and TransE-based methods in entity alignment. Moreover, the retrieval accuracy of key node similarity is better than that of global similarity, while reducing retrieval time by 19%. Building on the above findings, this work provides a novel perspective and approach for the efficient reuse of remanufacturing process knowledge.