Amazon reviews and tags don't always get used "seriously" — the Kevin Federline album Dan linked to is a great example. It's been somewhat of a Amazon tradition to target certain products with irreverent and funny reviews. Sometimes, they're removed by Amazon since they're pretty inappropriate, but others, such as reviews for David Hasselhoff's album, whole milk, or recreational tanks (…interesting cross-sells!) have been left alone.
Now, the point of writing these reviews is probably not to shill for the products and make more sales, but their existence has undoubtedly driven disproportionate amounts of traffic to the product pages since people find them funny and send the links to their friends. How does phenomena like this fit in with recommender system security? Is Amazon benefiting from it?
