|
|
|
|
|
 |
Friday, September 19, 2003 |
|
Am I missing something or is Amazon using background assumptions in their ratings now?
Out of curiosity and cuz I'll maybe get it for the kids, I looked up Madonna's new kid's book on Amazon.
Right now, it has a 3 star "average customer review" rating.
But as far as I can tell, as of this morning the customer ratings are all 4 or 5 stars.
One way of explaining explain the discrepency is to imagine that they are using a middling assumption of 2 or 3 stars which has to be overcome by a large number of customer ratings. Under this scenario there are more customer ratings, Amazon's "average customer review" would asymptotically approach the actual average (and due to rounding to the nearest half star, would apparently equal that average).
The actual computation could be a "formally correct" one based on Bayesian statistics to combine "background knowledge" with incoming real-world data, or a kluge that gets approximately the same effect.
Or is there something I'm missing here? If I'm correct, Amazon is now doing something that as far as I know, my company pioneered on Emergent Music. |
|
In my opinion, it's somewhat likelier than not that Apple has an ultra-secret internal team working on the Cocoa port of OpenOffice. They can't even tell the guys who are publicly working on it that this is going on, because as soon as Microsoft knows about it, MS Office will be in serious risk of being dropped, as was the outcome for IE when Safari was announced. If MS Office is dropped before Apple is ready, that would have potentially serious consequences on Mac sales.
Apple carried out that strategy with Safari. Like OpenOffice, Safari is based on an open source project, Konquerer's KHTML. Apple developed Safari internally with no word to the folks working on KHTML. When Apple was ready to make an announcement, they integrated their modifications into the main KHTML line. The people who had been working on it publicly were surprised but happy.
The exact same scenario could be underway with OpenOffice.
The strongest arguments against the above scenario are that Microsoft may have already made Apple agree not to support OO in return for MS's continuing to develop Office, or may have issued (possibly subtle) threats on the subject. But MS may be afraid to do this for fear of adding more evidence to their antitrust battles.
|
 |
 |
|
 |
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||