Big Five Personality Type Prediction on Twitter Users with the Long Short-Term Memory Algorithm

Every individual has a different personality. By knowing your personal personality, you can develop your potential to the fullest, by knowing the personalities of others you can position yourself socially. One of the individual personalities that is internationally recognized and recognized by psychology is the big five personality. This study aims to predict the big five personalities of individual Twitter social media users. Research activities follow the stages of the Cross Industry Standard Process for Data Mining (CRISP-DM) method, while the algorithm used to predict the big five personalities is Long Short-Term Memory (LSTM). The data used in this study is a public dataset published in previous research which contains 269 Twitter account user data with 240,000 tweets data. This study uses a variety of training data and test data as well as the Convolutional Neural Network (CNN) algorithm as a comparison of the research results. The results of this study are that the LSTM algorithm produces the best accuracy of 77.72% in the comparison of training data and test data of 80:20. These results are better than using the CNN algorithm and the random forest algorithm in previous studies with the same data splitting scenario dataset.