Research of Product Data Measurement Mode Based on Neural Network

With the evolution of traditional brick and mortar retail stores to online shopping, consumers are posting reviews directly on product pages in real time. We ultilize the RBF network to predict the sales number. We construct a GM (2,1) model to predict the sales of hair dryers, microwave ovens, and pacifiers in the next five years. The state transfer probability between positive, neural and negative reviews is built. The GINI correlation coefficient method is applied to determine whether the reviewer s attitude has a strong relationship with the star rating. The innovation of this article is that we build a comprehensive model with neural network, Markov and GINI correlation coefficient. Besides, Pearson correlation coefficient is used to test the GINI correlation coefficient which make the result more complete.