Rang a few bells when a few years ago we tried to use clustering the archives of some of Reed Elseviers publications , New Scientist (50 years) and Flight International (100+ years).
Id put together a few tests - but we couldn't get Anyone to take notice.