A multimodal dataset bridging brain activity, cognition, and real-world collective behavior

Abstract

Abstract We present a multimodal dataset integrating brain activity, cognition, and real-world collective behavior using naturalistic video stimuli recorded in a controlled laboratory setting. The dataset includes 64-channel electroencephalography (EEG) recordings from 65 participants during narrative video viewing, multidimensional subjective evaluations spanning affective, cognitive, and social domains, and participant-level demographic and psychological trait measures. The stimulus set consists of videos organized across seven semantic themes and contrasting narrative perspectives (grand vs. individual). Importantly, each stimulus is associated with real-world engagement metrics, including play counts, likes, comments, danmaku, saves, and sharing behaviors collected from an online video platform. The dataset also includes raw and preprocessed EEG data, stimulus annotations, and preprocessing scripts. This resource supports research on naturalistic information processing, neural synchrony, narrative-related cognition, affective computing, computational social science, and neuroforecasting, and enables investigation of how individual neural and psychological responses relate to population-level behavioral patterns.