A Forecasting Support System For Aquatic Products Price

The government, scientific research organization and enterprise always focus on aquatic products price, because which play an important role in development of fishery economic and international trade in China. That predicting the aquatic products price and overseeing its change not only can provide support for enterprise decision-making, and also allocate aquaculture resource more effectively and weaken its fluctuation. So we put forward to study on a forecasting support system for aquatic products price, which can provide information service for aquatic products marketers. We had fulfilled three aspects of the research.1) We had recognized the domain problem of aquatic products price forecasting support system and built the conceptual model and formalism architecture, which consist of 8 kinds of elements, 14 kinds of definition and 5 kinds of theorem and describe the domain problem. Building the model and architecture is useful to provide the support for design the system prototype and promote the reuse of system.2) We had defined the predict model for aquatic products price to be a 5-elements-set, which is the set of input and output, the condition, the structure and algorithm, and divided the procedure for building the forecasting model into 3 stages including 6 steps. Based on the procedure, we had constructed the composition model of time series and casual model for predicting aquatic products price.3) We had constructed the structure of casual model via the function of inverse demand and supply. But because of shortage of empirical information, we had corrected the structured via the theory of price transmit and price exception. Then we adopted the cointegration and error correction model as the algorithm of the model, which can resolve the line association in independent variable from inverse demand or supply model without losing the variable.4) We put forward and proved that the output structure also is one of the causes resulting in the price fluctuation. Then we analyzed the relation between output structure and output growth via Moore index. It is proved that the relation is positive association. But the result from cointegration show the structure of fresh water catch can't result hi the price fluctuation5) We had build all of the empirical models above applied the theory of economics, prediction and econometrics by farmer price and retail price of aquatic products in 2000 sampled between 1978 and 1999. That all of them is successful to test and evaluate shows that both the models and the thought constructed them are effective and feasible.6) We had designed and developed the prototype of the system.