Abstract
INTRODUCTION: Massive Open Online Courses (MOOCs) are gradually integrating GenAI technologies to provide personalized support. However, the mechanisms underlying the factors influencing learners' satisfaction in GenAI-supported MOOCs remain unclear. Under learning experience perspective, this study constructs a model of factors influencing college students' learning satisfaction (LS) that includes four dimensions: GenAI-supported MOOC learning environment (LE), teacher-student interaction (TSI), student-student interaction (SSI) and learning outcomes (LO). METHOD: Data from 402 college students' questionnaires were collected in GenAI-supported MOOC courses, which was analyzed by IBM SPSS 25 and AMOS 26 software platforms. A structural equation model (SEM) was used to validate the theoretical model of satisfaction influencing factors. RESULTS AND DISCUSSION: Results found that: (1) students' overall LS in GenAI-supported MOOCs is high, indicating the environment can satisfy students' most learning needs; (2) although LE does not directly affect LS, it positively influences LS through the mediation of LO, which suggests that learners' perception of LE needs to be translated into the actual LO before it can improve LS; (3) TSI has a significant positive impact on LO, but a negative impact on satisfaction, indicating that GenAI intervention may lead to emotional detachment or excessive expectations; (4) SSI promotes LO, but does not have a significant impact on LS, reflecting that the value of peer collaboration has not been fully embodied in GenAI environment. In GenAI-supported MOOCs, improving teachers' GenAI collaboration ability, balancing human-computer roles, and strengthening emotional support are the future directions to enhance LS. This study provides empirical evidence for the in-depth application and effect enhancement of GenAI in MOOCs.