Recent advances in natural language processing (NLP) have provided psychology with novel methodological tools. Language, as a key carrier of human psychology and behavior, enables the automated revelation of psychological characteristics across individuals and groups, spanning dimensions such as emotion, cognition, personality, social relationships, and culture. This article systematically reviews the latest developments in the application of NLP within psychological research. It focuses on its use in psychological construct mining, multilingual text analysis, automatic scale item generation, machine-assisted hypothesis generation in social psychology, group-level psychometrics, and the prediction of mental health intervention outcomes. Research indicates that NLP methods based on large language models (LLMs, e.g., GPT, BERT) not only enhance the efficiency and objectivity of psychological measurement but also expand the diversity and ecological validity of research samples. However, these approaches still face challenges related to data representativeness, model bias, interpretability, and ethical privacy concerns. Future research should deepen interdisciplinary collaboration between psychology and computer science and advance the development of interpretable NLP techniques and cross-culturally adaptive models.
Jiale Lv (Thu,) studied this question.