It could be interesting to further enrich the data using info from urlscan.io or something similar.