The Research Of Comparison Shopping Model Based On The Ant Colony Optimization Neural Network

Base on the research on the majority of existing comparison-shopping Web site, this paper puts forward a new model of comparison shopping. On the basis of the traditional comparison shopping, learn for the user's preference information which based on the ant colony optimization neural network, and thus to re-sort search results, to provide users with the personalized shopping guide service which meet their needs.This paper first describes the creation and development of comparison shopping, discussed and put forward the most current comparison shopping site problems:only focus on consumer demand on commodity prices, while ignoring the other aspects of consumers about the product requirements. Meanwhile, this paper studied the basic principles of neural network and ant colony algorithm, as well as their development application. After analyzing the basic characteristics of these two technologies, this paper put forward a method of combining the two technologies, through the ant colony algorithm to optimize the BP neural network, the method can improve the efficiency of comparison shopping search engine and be able to avoid the BP algorithm defects.Base on the Combination of two algorithms, this paper is completed the model design of comparison shopping which based on the ant colony optimization neural network, and related features of this model and the specific processes in detail. Finally using MATLAB software for the simulation experiment, Experimental results show that ACO-BP neural network in comparison shopping search model is effective, user's shopping preferences can be well simulated according to the training samples, so as to make comparison shopping search guide. Finally, we compared the ACO-BP algorithm and BP algorithm in the performance differences, confirmed the ACO-BP algorithm can effectively overcome the shortcomings of slow convergence of the BP algorithm.