What is the scheduling algorithm?

[crgm...@yahoo.com.au]

Hi everyone,

I must confess, I am not actually a user of this product, but I think
it sounds like a good idea and it's great that it's been made
available as freeware. I actually have my own home-made software which
does something vaguely similar (though it's currently narrowly
targeted at learners of German.)

I joined the group because I have been thinking about scheduling
algorithms for quite some time, and then stumbled across mnemosyne and
saw all the same issues being discussed.

Given that this seems to be an open-source project, would it be
possible to see the algorithm, either in plain English or, better, in
pseudocode? Many of the posts to this group show that users don't
always have a good sense of what is going on under the hood, and the
explanations from the authors are often impressionistic, rather then
precise.

Some day when I'm less busy I might be able to contribute as a
developer, but for now I'd be very interested to see the core of the
algorithm, or at least get an idea of what the default scheduling
intervals are, and how sensitive the algorithm is to the past history
of a card, and how its various previous exposures and their gradings
contribute to the final interval. Is that possible? If you'd prefer it
to be semi-secret, that's fine too.

Another aspect of the same question is, how many global adjustable
parameters are there in the algorithm? If there are only a few, the
program could search the relevant parameter space by 1) adding a bit
of judicious randomisation at the time a card is born, so that the
card has its own unique variant of the algorithm, and then 2) keeping
tabs on the card's eventual journey to full mastery, and finally 3)
comparing the success rate and acquisition speed across the variants.
Or do you already do something like that?

Sorry if this has been asked before, or is posted somewhere else. If
so, please just point me in the right direction.

Cheers,

Craig.

[David Alt]

You'll find detailed information at the supermemo site, where it was
initially developed into software.
http://www.supermemo.com/english/algsm11.htm

Note there are variations on the algorithm already; in addition to the
natural evolution by Supermemo project, and any that may have occurred
at Mnemosyne, I know Anki has introduced its own variations on the
algorithm as well.

[Peter Bienstman]

Actually, it's SM2, not SM11

http://www.supermemo.com/english/ol/sm2.htm

Peter

--
------------------------------------------------
Peter Bienstman
Ghent University, Dept. of Information Technology
Sint-Pietersnieuwstraat 41, B-9000 Gent, Belgium
tel: +32 9 264 34 46, fax: +32 9 264 35 93
WWW: http://photonics.intec.UGent.be
email: Peter.B...@UGent.be
------------------------------------------------