SUMMARY
Personality classification is one of the ways in the field of Natural Language Processing (NLP) with a collection of data that can describe the user's personality through input sets of text documents such as status uploads. Social media is one way to interact online that can provide convenience for users, such as interacting, expressing themselves, and expanding friendships. Status posts on social media can be extracted into useful information in the personality classification process. This research performs classification based on how social media users live their lives and make decisions, which is a representation of the "Thinkers/Feelers" and "Judgers/Perceivers" class attributes of the Myers-Briggs Type Indicator (MBTI) model. Researchers are encouraged to develop a personality classification system with feature extraction that can improve system performance. In this research, there are three main experiments conducted, experiments using data with oversample techniques in the Thinker/Feelers (TF) and Judgers/Perceivers (JP) classes provide the best results compared to other experiments with f1-score and accuracy of 92% using the Random Forest classification method and Glove as the extraction feature.