Under the background of big data, the complexity of industrial products gradually increases. It is necessary to study efficient production methods and use big data-driven methods to improve the productivity of complex products. Based on modularization, this paper focuses on the module partition problem of complex products and transforms this problem into the community detection problem of multi-relational networks. First, the multivariate relationships between components are mined based on the data of the complex product. Then, this paper innovatively considers the supporting relationships to demand and the dependency relationships on technologies of components and proposes the co-supporting and co-dependency correlations of components. Combined with the structural and functional correlations of components, a multi-relational network of components is constructed. Taking modularity as the optimization objective, the optimization model is built for the partition of the complex product. Next, a heuristic module partition algorithm based on the improved Genetic Algorithm is proposed to realize the module partition of complex products efficiently. Finally, a case study is carried out to expound utility and effectiveness of the proposed methodology. This research is of great significance to the modularization of complex products.