Amazon Patents User Viewing Histories

Chris Cleveland writes "Yet another astounding patent from the USPTO. I was browsing the patent database, and discovered that Amazon received a patent today on using customer viewing histories to generate recommendations. If a customer views product A, and then later views product B, and you use that to infer a relationship between A and B, then you've infringed on this patent. This patent is a continuation of an earlier patent (#6,317,722) on using shopping carts to generate recommendations. When will this stupidity end?"

Comments

Here's how you end this madness

[melted]

Go ahead and patent the patent process. Then impose heavy licensing fees, a million dollars per patent should be enough.

Vagueness vs Infringement

[NetSettler]

Is there someone reading this discussion that knows what the status of infringement law is when it comes to vague claims like this?

For example, suppose I build a system that uses past viewing histories AND something else. Have I infringed, or have I used a different technique?

What would happen if I patented "using the letter E in the making of recommendations"? Would that mean that anyone who uses the letter E and also does other things to make recommendations is infringing? Or would it mean they are doing "something else"?

It seems to me trivial/obvious that you use "past knowledge" to make recommendations, so I'd be surprised the patent was granted if I didn't know that pretty much all patents like that get granted these days. But beyond the question of obviousness, it's equally obvious that this alone can't really give you the best solution. So am I "doing something else" or "refining this idea" if I use both this info and other info?

Or is that something courts disagree about?