Intelligent Supplier Matching Method Based On Semantic Graph Model

The age of intelligent manufacturing has arrived,the deployment of supply chains has become increasingly global and quick,and it is very important for companies to have an efficient method for quickly evaluating and selecting suppliers.In the past,people have done some research in this field.Because the supplier selection problem is affected by various factors,the effect of applying supplier matching method to large datasets is not ideal.The existing research has the following problems: first,the information of supplier matching is inconsistent,and the formal representation of supply and demand relationship is not perfect;Second,there are many similar products and functions,so it is difficult to deal with products and functions with similar concepts;Third,supply chain deployment lacks speed and globalization,and efficiency and accuracy need to be improved in the face of big data sets;Fourth,the complexity of user operations is high,such as the weight depends on the user specified.In order to face these challenges,this paper has done the following research:(1)In order to solve the problem of information inconsistency in supplier matching,this paper defines a supplier description model with universality and semantic information,converts supply and demand into graphs according to the model,and then solves the information inconsistency and avoids the possibility of semantic information misreading.(2)For the problem of similar products and multiple functions,this paper defines the domain concept graph,classifies the concepts of products,and designs the calculation method of similarity between concepts.Combining with the threshold value of similarity of concepts,the concept products with low similarity are discarded.(3)Considering the deployment of the supply chain quickly and globalization,this paper designs the four stages of suppliers matching algorithm based on semantic graph model,the graph matching algorithm is used to calculate similarity between demand and supply graph,contains the calculation of similarity between nodes,the calculation of weight and the weight distribution,the calculation of similarity between supply and demand graph,reduce the complexity of the user action,finally,a ranking set based on similarity is obtained in a large number of supplier resources,which is used for supply chain deployment recommendation.Users only need to input product requirements and enterprise requirements,and then they can get a matching set of suppliers to implement the idea of intelligence and convenience.(4)The subgraph matching problem is recognized as an NP-complete problem.This paper proposes to add the "entry node of the graph" to the maximum subgraph matching problem as the starting position of the graph matching process to reduce the search space and use the entry node to guide the subgraph matching process.The graph matching process is made close to the tree traversal process from the root node,so that the subgraph matching problem is completed in polynomial time,and the time complexity is reduced.