Study On Shopping Online Behavior Diffusion Forecasting Model Based On Bass Model

The rapid development and popularity of online shopping provides unlimited business opportunity, but it also brings large risk. Facing the field containing business opportunity and risk, companies have many question, such as the estimation of market potential, the decision of input quantity ans so on, the questions is important to company’s future development. Company will obtain the competition advandage after it gets the answers above.In recent years, more and more scholars begin to cares about the online shopping diffusion forecasting problems, and use Bass Model which being used in entity product to forecast online shopping diffusion. Becase online shopping’s carrier is information and technology, it is different from entity product’s carrier, so it have to prove the fitness and validity about whether Bass model can be use to forecast online shopping’s diffusion.Given the online shopping forecasting needs as well as the lackness of theory study, this thesis studies online shopping based on the Bass model, expand the following aspects:(1) We check the fitness of Bass Model to online shopping behavior diffusion. This thesis proves the problem from three views as a progressive way. The views are basic principle, model building key factor and data requirement. First we analysis the fitness from basic principle view, we find that the diffusion process and innovation charactors satisfy the basic priciples; second, we analysis the fitness from Bass Model key factors views, we find that online shopping behavior diffusion is similar to entity product, both of which have the three Bass Model key factor. Third, we analysis the fitness from model data view, we find that online shopping behavior innovation data fit the data series and model fitness require. At last, we empirical the model with the online shopping behavior diffusion data, the result shows that the fitting degree and forecasting accuracy all pass the statistical test.(2) We put foreward dynamic market potential online shopping prediction model. Online shopping only diffuse in the internet users just as the common sense tells us, the upper limit of online shopping groups is the number of Internet users. As the number of Internet users change over time, whereby the proliferation of online shopping market potential is non-fixed, but varies with the increase of Internet users. Based on the above characteristics, we replace the fixed value of market potential in the Bass model with a function of time about Internet users to express to build online shopping dynamic market potential forecast model. The model includes not only the proliferation of online shopping information, but also contains the proliferation of the number of Internet users which will ensure that the forecasting result is more realistic. Comparing the result of Bass model and the market dynamics potential model, the latter is more accurate and precise.(3) We establish the word of mice online shopping model. Through comparative analysis of the word of mice, the mass media as well as traditional word of mouth, we can get that word of mice is different from the other two. On the basis of the original two diffusion channels of Bass model, we add Word of Mice information channel into the model. We sort the influence of the three information channels based on estimation result. The empirical results show that the word of mice improved model’s fittness and accuracy are improved better.(4) We found online shopping system forecasting equations. First, we use content analysis method to determine the relevant factors of innovation within the online shopping system; the analysis showes that the three most relevant innovation factors are:cash flow, information flow, and logistics. Then we establish online shopping system forecasting equations, the equations only contain online shopping information, but also include capital flow, information flow, logistics diffusion information. The forecasting equations contain more information, are more realistic and avoid the incomplete information problem in single innovation diffusion effectively, increase predicting accuracy. The system forecast model provides online shopping forecast information, it is also able to provide other innovative forecasting information within the system, provide the basis to business and government for their decision making.This thesis studies online shopping behavior diffusion, and make progress on the three key factors:adoptor maket potential, inner effect and outer effect. This paper gets some innovation points as below:first we puts forward that adoptor maket potential is dramatic and this paper uses dramatic function replacing the original fixed constant, provide effective tools to estimating online shopping diffusion potential; second, this paper inputs word of mice to Bass Model, make to model contain the effect of word of mice; third, this paper draw lessons from coordination theory, and found online shopping behavior diffusion forecasting model, this paper considers not only the online shopping itself, but also other coordinating factors, so to provide wider studying views to other scholars.