Dataset for the LREC-COLING 2024 paper "A Gaze-grounded Visual Question Answering Dataset for Clarifying Ambiguous Japanese Questions"