On Wed, Jul 18, 2012 at 2:25 PM, James A. Donald <[email protected]> wrote: > > What we want is peers that are trusted by entities like ourselves, and/or > have engaged in transactions that are beneficial to entities like > ourselves, not those that allegedly trust entities that we trust and have > allegedly engaged in transactions like those that we have engaged in.
The dataset I wish to collect and make public doesn't say anything about what the transactions actually are. The optimizing factors are success rate and transfer rate. Slope One and Singular Value Decomposition does not trivially give us > that, and it is not immediately obvious to me how to fix them to give us > that. I expect it can be done, just do not quite see how. These algorithms are typically employed for "recommendation systems" such as the one seen on Amazon, i.e. "based on your behavior we think you'll like products X, Y, and Z", where recommendations are driven by a large corpus of user data. I am attempting to perform a similar calculation, except in this case it's "based on my behavior I think I'll like peers X, Y, and Z", and the calculation is driven by a large corpus of peer interaction metadata the system collects and distributes by design. -- Tony Arcieri
_______________________________________________ p2p-hackers mailing list [email protected] http://lists.zooko.com/mailman/listinfo/p2p-hackers
