Understanding Chinese university students' continued learning intentions with DeepSeek: An Integrated TAM-SOR model

BACKGROUND: DeepSeek has the potential to drive the modernisation of higher education in China. As the primary beneficiaries of higher education, university students are transitioning from traditional learning methods to intelligent learning methods. However, current research in the field of educational artificial intelligence remains excessively focused on rational dimensions such as knowledge transmission, intelligent assessment, and personalised learning, while significantly neglecting emotional perception and sustained learning motivation - factors that are crucial to the learning process of university students. METHOD: We combined the Technology Acceptance Model with the Stimulus-Organism-Response model to analyse the emotional perceptions influencing their use of DeepSeek as a learning tool. This study collected 476 valid responses through an online questionnaire survey. CONCLUSIONS: This study reveals the pivotal role played by perceived emotions and sustained motivation in university students' adoption of generative artificial intelligence learning technologies. The findings indicate that students' perceptions of tool convenience not only enhance evaluations of practicality and usability but also stimulate positive emotional experiences-designs aligned with learning habits enhance a sense of control and pleasure, becoming the core drivers for sustained use. Students with high self-efficacy demonstrate unique advantages, as they can establish an emotional connection between the tool's functions and their personal needs, forming the belief that 'I can use the technology effectively,' thereby effectively reducing usage barriers. This helps to understand the key factors influencing university students' sustained adoption of learning technologies and reduces perceived barriers to the adoption of generative AI learning platforms among university students.