In order to meet further understanding of the financial market and the demand for the stock price forecast in advance. This paper proposed a stock price forecasting model based on TA/SA (technical analysis/sentimental analysis), which aims to establish the relationship between sentiment and future stock price. The application mainly includes: get sentiment index, establish regression model and the calculation of future stock closing price. This model predicted 200 stock prices, compared with predicted results of SVM and BP neural network, and the correct rate of the proposed model increased by 10.9% and 7.4%, which shows that the model has better prediction accuracy and practical value.