MBTI Prediction Study Using Word2Vec

As internet usage increases, the amount of data, such as text and images, being created is growing dramatically. Nowadays, we live our lives by sharing and obtaining information through various online platforms. Since individuals primarily share their lives and information through text, text not only contains information but also conveys emotions and psychological states. Additionally, personality type analyses like MBTI are used to understand an individual's psychological state. However, MBTI requires a significant number of surveys and is challenging to capture rapidly changing emotions. This study utilizes Twitter-based data to train an MBTI prediction model and analyze the personalities revealed in the results through short texts. Furthermore, by analyzing written content, it can determine current emotional states and identify words and sentence patterns predominantly used by each MBTI type. Through this research, rapid psychological analysis can be conducted using short texts, and user patterns can be predicted. This study can check while exchanging conversations or texts, and reduce the cost of psychological and personality tests.