I know, wishful thinking. A little scraping might go a long way. If the data were available, I'm certain the networks could be easily identified by clustering t

I know, wishful thinking. A little scraping might go a long way. If the data were available, I'm certain the networks could be easily identified by clustering the overlap on 5 star reviews of totally unrelated items. e.g. I noticed a highly unlikely overlap between multiple reviewers of a particular hardware product my wife had mistakenly bought. When reviewing how she could have made this mistake, I discovered the title was misleading and the many 5 star reviews were fake. Looking at 5 star reviewers history showed a bunch who shared reviews of completely different items (in the make-up/food supplement arena) all produced by the same manufacturers.