Haven given a lot of thought as to which site would be the best fit for this question, I decided it was ultimately a user experience one. Feel free to suggest a better site if you think it would fit better elsewhere.
The paper, From Amateurs to Connoisseurs: Modeling the Evolution of User Expertise through Online Reviews heavily focuses on data mining and customer reviews. It's main argument is that users with a similar level of experience on a subject, tend to give similar ratings. For example, someone who is new to drinking alcohol may not like whiskey if they've never had alcohol before, but after growing accustom to it, they may find very enjoyable.
This made me wonder, what is the purpose of ratings? Why is it Amazon has 5 stars that customers can choose and IMDB has 10 stars a person can rate. For example both Citizen Cane and Starwars the force awakens have 8.4/10 stars on IMDB. What information is this actually trying to convey to the user? It seems like comparing apples to oranges to say that means they are exactly the same level of "goodness".
On Amazon it wouldn't make sense to conclude that because a toaster had 4/5 stars and a hairdryer had 3/5, then it's better to get the toaster.
In conclusion, I am concerned that such rating systems lead the way to people trying to force a statistical pattern, where none actually exists.