Last week I posted 3 plots comparing 2 tags each:
This page is one step closer to my end goal. I've 10 tags and 500 books and mapped the first 2 principal componets. The tags I've mapped are:
This isn't a spectacular cross section of the 5000+ tags I hope to work with in the future. Also the 500 books aren't the perfect representative sample of the chosen tags.
Even within those constraints, you can still see some patterns to the data.