An Electroencephalogram Dataset of Learner Interest States in Online Education Tasks

Learning interest is widely recognized as a critical factor influencing learning outcomes; however, its underlying neural mechanisms remain insufficiently understood. In this study, we constructed an EEG dataset within an online learning context to explore the neural representation of learning interest. Fourteen participants were recruited to complete comprehensive online learning tasks in simulated multi-context classroom settings, during which 16-channel EEG signals were recorded synchronously. The dataset comprises EEG recordings collected during the learning process together with the corresponding video materials. To validate its effectiveness, five machine learning models were applied to classify interest states based on spectral energy and peak-frequency features extracted from five canonical EEG bands. Classification accuracy approached 100%, confirming the reliability and analytic value of the dataset. This resource provides empirical support for investigating the neurophysiological mechanisms underlying learning interest and, as an open-access dataset with ecological validity, offers valuable data for research in online educational neuroscience, cognitive learning processes, and brain–computer interface applications.