In order to solve the problem of prediction for long-term web sentiment, a prediction model is built using the proposed method in this paper. First, a novel clustering method based on the extension neural network (ENN) is introduced to recognize the types of subclass of web sentiment. For each class of social events, the class model library of the development trend of web sentiment is established by cycle analysis and ENN clustering combined with nonparametric auto-regression analysis (NAR) method. Then the adaptive transformation is applied to the already known development trend of a new social event, and the min-sum of mean square error (MSE) from the library is selected to predict the future development trend of web sentiment. Empirical findings indicated that compared with the traditional methods, such as the GM (1,1) and least squares estimation (LS) method, the approach presented in this paper yields a higher correlation value in predicting the long-term development trend of web sentiment and can predict the turning points of the development trend more effectively. The ENN- and NAR-based prediction model can effectively solve the problem of prediction for long-term web sentiment.
展开▼