Abstract Audiences’ reviews are critical to the reception study of movies. This article takes the reviews of Hero as an example, and analyzes the reception effect from a digital-intelligent humanities perspective with transformer-based models, especially bidirectional encoder representations from transformers (BERT)-based sentiment analysis and BERTopic modeling, which are generally regarded as the state-of-the-art deep learning models. The results of sentiment analysis show that positive comments of Hero account for 74.42 per cent, while the proportion of negative comments is 25.58 per cent. Besides, the Maslow’s hierarchy of needs theory is employed to further reveal the actual feelings and needs of audiences, including physiological needs (visual and auditory needs), safety needs, social needs (discussion on plots and cultural sharing), esteem needs (themes and emotional expressions), and self-actualization needs (cultural participation, aesthetics, and knowledge extension).
Yiyi Hu (Mon,) studied this question.