Modeling The Diffusion Of Innovation Base On BA Scale-Free Network

The relationship between social group formed a complex network, The diffusion of new product in the crowd is comparable to spread in a complex network. So the structure of network can cause momentous impact to the diffusion of new products. But in the field of forecasting, one of the most important method in forecasting the diffusion of innovation------Technical innovation, diffusion models, cannot combined the structure of network with the models reasonably:Macro diffusion models completely ignore the micro mechanism of adoption, which includes the structure of consumer network, consumer behavior and so on. Micro diffusion models simulate diffusion process mechanically base on network structure, and it cannot be empirical vilified. So, how to consider the micro mechanism of diffusion logically, and how to combine network structure with macro diffusion models, are essentials to enhance the predict accuracy of innovation diffusion models.Therefore, by analyzing the merits and disadvantages of various technical innovation diffusion models, and the simulate results of various complex networks, this article modeling the diffusion of based on BA scale-free network. And we use the actual data for model validation. This paper focuses on the four tasks which are shown as follows:1,Summarizing innovation theory and innovative models. Analyze the directions to improve technology innovation diffusion models. 2,Analyze the basic theory and structure of complex network, provides theories and methods to the establishment of the diffusion model based on BA scale-free network.3,Based on the analysis of telecom products'market potential in China, we set market potential expression of telecom products. And then we establish BA scale-free network diffusion model by combining the market potential expression with the increasing of network nodes in BA scale-free network.4,Using the data of mobile phone subscribers and broadband subscribers, we analyze the goodness-of-fit of BA diffusion model, Bass model and SWN model to prove the effectiveness of the BA diffusion model; And we use the BA diffusion model to forecasting China'mobile phone subscribers and broadband subscribers in the future.