Paper Title
Emotion Prediction using an Efficient Psychological Question Mining Method

Abstract
Emotion questionnaire is becoming increasingly important for measuring psychological emotions. They are designed to measure in categories to which individuals appraise situations in their lives as emotion categories. The using of long psychological questionnaires in this context may be exhaustive under an emotion situation because it is important for participations to bein screening. In this study, we propose an efficient method for designing predictors of psychological emotion using an optimal set of psychological question items obtained by using an inheritable bi-objective genetic algorithm (IBCGA) from our dataset which is composed by71 adults with anger, 96 adults with social anxiety and 144 adults with inferiority who responses to the emotion category of 77 questions. For the independent-adult prediction, the training dataset contains 80% adult with three types of emotions and other serves as the test dataset. A predictor, PreEmotion, composed of 27 optimal features selected using IBCGA bases on an intelligent genetic algorithm is created for predicting three types of emotions with 96.53 % training accuracy and the test accuracy is 80.66%. Keywords - Emotion Questionnaire, Psychological Emotion, Genetic Algorithm, Prediction, SVM.