A Part Manufacturing Process Knowledge Recommendation Method Based on Hybrid Model

Process designing for complex parts requires a lot of process knowledge for reference, but if the user needs to retrieve process knowledge during process design will take a lot of time, so the user process knowledge recommendation is a better way, but it is difficult to guarantee the recommended precision. This paper proposed a process knowledge recommendation method based on a hybrid model. Firstly, the user's knowledge requirements model is constructed based on the user's role and task process knowledge category to calculate the user's requirements for process knowledge. At the same time, the matrix decomposition model is used to fit and predict the user knowledge scoring matrix generated by the user's historical behavior, and the user-knowledge scores prediction matrix is obtained. Finally, the user-knowledge requirements matrix and the user-knowledge scores prediction matrix is fused by dynamic weight calculation, and the knowledge items are recommended to users according to the scores. Finally, an experiment proves the effectiveness of this method. Experiments show that this method has higher precision than the traditional collaborative filtering method, which will provide meaningful help for process designing.