g-Rel-READER

Information relevance judgment is a fundamental cognitive decision performed by users during information seeking and reading tasks. Inferring relevance decisions from neurophysiological signals, particularly through eye-tracking and EEG recordings, opens possibilities for real-time relevance detection. To support this line of research, we present g-Rel-READER, a unique dataset comprising eye-tracking and EEG recordings collected during reading and relevance judgment of short news stories in a question-answering task. The dataset provides screen coordinates for each word, enabling temporal and spatial alignment between eye-tracking data, EEG signals, and the specific words being read. Question-relevant words are explicitly marked, facilitating analysis of relevance judgment processes.