The Impact of Ancient Book Image Quality on Cultural Heritage Crowdsourcing Correction Tasks: A Human Error Theory Perspective

ABSTRACT This study investigates the impact of ancient book image quality on the accuracy of cultural heritage crowdsourcing correction tasks. The study processed 27 pictures from an ancient book to adjust their quality in terms of clarity, seal coverage and background color. Based on human error theory, the study divided common errors made by participants in correction tasks into slips and mistakes, and analyzed the impact of different image qualities on these errors. The experiment was conducted using an online experiment platform, with 25 participants each completing 6 tasks related to clarity, seal coverage, and background color. The results show that the clarity of ancient book images significantly affects the number of rectifications and mistakes in the tasks. Different degrees of seal coverage significantly impact the number of rectifications and slips, while the background color of the images mainly affects the number of mistakes.