Research And Application For Data Mining In The Prediction Of Passenger Traffic Of Expo And Social Media

Along with the development of computer and storage technology, the ability of collecting data ability is greatly improved. Because of explosive growth of data, the ability of excavating useful knowledge from mass data is becoming more and more important. The World Expo 2010 Shanghai China was the most expensive Expo in the history of the world's fairs, having the most visitors and countries and international organizations having participated. During 184 days, we had recorded a lot of actual data, like video data, passenger flow data every 5 minutes and ticket sales data, etc. How to explore some useful knowledge from collected data to predict future , for example, predict the passenger flow every day, certainly will help world expo organization schedule human, financial and material resources to ensure safe and effective operation of the world expo. Predicting the passenger flow accurately is significant, thus, in this paper, forecasting the passenger flow is our first study point. On the other hand, the rapid development and popularity of the Internet and social media (like Weibo, Renren web, etc.), produces a large number of data whose notable characteristic is that they timely, accurately reflect the true feelings of users. At now, exploring some useful knowledge from these data to forecast the future is the hot and difficult studying issue. At the same time, the enterprise brand all post their own new products through the social media to improve awareness of their products amongst the public. Naturally, discovering and exploring knowledge through data mining on social media data to predict the product's future performance may help enterprise decision-making. It is great significance. Based on the successful prediction of the passenger flow, we will use the most popular social media-Weibo-data, to study the movie box office forecast method.The passenger flow has the very strong randomicity and nonlinear, while artificial neural network is a nonlinear dynamic system, which can realize the nonlinear mapping of variables in arbitrary precision, and have good flexibility, learning ability and generalization ability. The characteristics of the neural network, to a certain extent, meet the demand of forecasting. At the same time the passenger flow data has its special features that the peak of visitors appeared between 10am and 12am, and the linear relationship between the peak point and the number of visitors per day is strong. Therefore, finding out the inflection point, so as to forecast the linear part of the passenger flow forecast, has vital significance. In this paper will put forward the algorithm calculating inflection point and establish inflection prediction model of the passenger flow. And we will also combine linear forecast model and nonlinear prediction model, and then put forward the combination forecast model to overcome some faults of a single prediction model.In the study of using social media(Weibo)to make the forecast of movie box office, we found that changes in the rate of Weibo is able to reflect the trend of the movie box office, and it exists strong linear relation between them. We use the Weibo rate to establish linear regression model, and the experimental results show that the model is effective and feasible. At the same time, we use data mining technology—emotional analysis technology—to analyze Weibo content, and then use the emotional value as an additional variable of regression model to establish forecasting model, and improve the prediction accuracy. We draw the conclusion: mine social media for knowledge to make prediction of the future is practical and feasible.The innovation points of this paper is as follows:(1) Put forward the inflection prediction model and combined the BP neural network with the inflection prediction model to forecast the passenger flow for Expo.(2) Did some data mining on the most popular social media-Weibo data, and established a linear regression model based on Weibo rate to forecast movie box office. (3) Studied emotional analysis algorithm to do emotional analysis on Weibo content, and make use of emotional value to predict the movie box office, improving the prediction accuracy.