A study on the discriminant accuracy of the trip purpose based on the volume of personal attribute information

The decision tree analysis for the trip purpose using time and space interpolated person trip survey data

Abstract

Recently, applications of the person flow analysis for the recommendation got lots of attention. A person flow reflects the circulation, location of individuals and travel history based on individual interests. Prediction of the person trip purpose enables more effective recommendation. For the accurate prediction, personal attribute information is essential.
In this study, we carried out the decision tree analysis for the prediction of trip purpose based on the characteristics of person trip and individual attribution, using time and space interpolated data of person trip survey. Accuracy of the decision tree analysis was improved according to the volume of personal attribute information (e.g. gender, age, occupation).