Re: Fwd: [Add a Category] The usefulness of Mnemosyne
Peter Bienstman
(I took the liberty of copying this message to our discussion forum at
http://groups.google.com/group/mnemosyne-proj-users?hl=en, where perhaps more
people can chime in with comments.)
As for your poem question, the very first showing of new cards can either be in
the order that they were added, or in random order. There is an option for
that in the configuration screen.
I personally would not memorise poems using Mnemosyne, as the main reason for
memorising a poem would be able to recite it from front to back, and not being
able to say line 21 after you see line 20. However, I know some people who use
Mnemosyne to memorise poems, and they do seem to like it.
You are right that Mmemosyne (and other similar programs) are mainly useful
for learning factual information, and not skills. However, it can supplement
the learning of skills if there is a certain amount of memorisation required.
So for your specific case of bridge, I would keep on improving your bridge
skills through playing, reading books, or whatever, and once you've got the
insight to distill a little 'rule' from that (like when it's OK to open with 3
clubs), you add it to Mnemosyne to make sure that you don't forget.
Distilling this set of rules would indeed be a mammoth task, but so would
learning to play bridge in the more traditional way, I guess. Once you create
such a deck, you could in theory share it with other people, but they probably
will not be able to pick up the 'insight' you got from playing by just
memorising the rule without context.
There are plenty of premade decks on our site, but I think for the learning
process (especially the 'insight' phase) it's always a good idea to create
your own deck from scratch.
The only exception might be premade decks which cover vocabulary from a
specific course, which is useful as a starting point.
Hopefully this answers some of your questions!
Thanks for the feedback,
Peter
On Wednesday, August 11, 2010 06:24:41 am Patrick Kenny wrote:
> On Tue, 10 Aug 2010 10:45 +0000, robin....@bookfeller.co.uk wrote:
> > Robin sent a message using the contact form at
> > http://www.mnemosyne-proj.org/contact.
> >
> > This is feedback.
> >
> > I was attracted to this application by an article in the press:
> > http://www.guardian.co.uk/lifeandstyle/2010/aug/07/change-your-life-mem
> > ory-burkeman
> >
> > (This article is about SuperMemo, but Wikipedia led me to your site as an
> > alternative.)
> >
> > My problem is, after trying your program out for a short while, that I
> > can
> > only see it as useful for memorising lists of facts (e.g. the dates of US
> > Presidents). WHat I had hoped was that it might be useful for memorising
> > two
> > categories of material in my own life: poems, and 'useful tips for
> > playing
> > contract bridge'.
> >
> > For the first use I tried breaking a long poem into couplets and putting
> > two
> > lines in as a question, and the next two lines as the answer. By linking
> > pairs of lines over a long set of questions I hoped would get help with
> > memorising the whole poem.
> >
> > The obstacle to this is that your card decks cannot (so far as I can see)
> > be
> > ordered. SO the fragments of poem come at you in haphazard order and the
> > thread (the structure) of the poet's thought is lost.
> >
> > I can't see a way round this. Putting the poem's first line as a
> > question
> > and the rest of the poem as an answer would just be silly.
> >
> > Am I missing something?
> >
> > On the second kind of use - learning to play better bridge - things seem
> > even
> > more problematic. What I had envisioned was putting in a series of
> > questions
> > like "When is it OK to open with three clubs?" and answering with some
> > short
> > notes about how many points you need, the balance of the hand, etc. But
> > this
> > would be a mammoth task. Getting a good set of questions and answers
> > would
> > require me to already have precisely that understanding of bridge play
> > that I
> > am trying to develop.
> >
> > SO, two thoughts:
> >
> > 1) AM I missing the point? Does you tool have features I have missed or
> > is
> > it really only suitable for memorising long lists of facts? (I admit
> > that in
> > exam situations - in some study areas - this would be a very useful
> > tool.)
> >
> > 2) Will your product only really take off when there are large decks of
> > cards, prepared by skilled experts, covering a range of topics? However
> > clever and well-founded the system is, how do you reach a critical mass
> > of
> > content as opposed to mere method?
> >
> > In friendship,
> > Robin Wilson.
Gwern Branwen
<Peter.B...@ugent.be> wrote:
> I personally would not memorise poems using Mnemosyne, as the main reason for
> memorising a poem would be able to recite it from front to back, and not being
> able to say line 21 after you see line 20. However, I know some people who use
> Mnemosyne to memorise poems, and they do seem to like it.
That'd be me, I take it. Robin, you can see my approach at
http://groups.google.com/group/mnemosyne-proj-users/browse_frm/thread/f69d567cc7608f52/84ce1bed201f6adc
Attached is a screenshot of a _Rubaiyat_ card in action. (You'll see
that it's a bit improved over the original; the latest versions of my
scripts can always be found at
http://community.haskell.org/~gwern/wiki/haskell/ )
>> > For the first use I tried breaking a long poem into couplets and putting
>> > two
>> > lines in as a question, and the next two lines as the answer. By linking
>> > pairs of lines over a long set of questions I hoped would get help with
>> > memorising the whole poem.
The combinatorial explosion of my approach means that it may not work
well for 6-lines or more verse forms; in that case, you might to try
memo2.hs. I don't seem to have written it up anywhere, but it's
basically taking it 2 lines at a time as you do.
> So for your specific case of bridge, I would keep on improving your bridge
> skills through playing, reading books, or whatever, and once you've got the
> insight to distill a little 'rule' from that (like when it's OK to open with 3
> clubs), you add it to Mnemosyne to make sure that you don't forget.
>
> Distilling this set of rules would indeed be a mammoth task, but so would
> learning to play bridge in the more traditional way, I guess. Once you create
> such a deck, you could in theory share it with other people, but they probably
> will not be able to pick up the 'insight' you got from playing by just
> memorising the rule without context.
I'll mention again that this is another rule-data tradeoff, where a
program-card (that generates random bridge situations according to a
rule and then tests you) would be useful.
See:
-
http://www.reddit.com/r/cogsci/comments/9aufn/ever_wanted_to_analyze_860mb_of_spaced_repetition/c0reso2
- http://groups.google.com/group/mnemosyne-proj-users/browse_frm/thread/b104b9a1543ba34/eb28b030b361f6b8?lnk=gst&q=multiplication#eb28b030b361f6b8
- http://groups.google.com/group/mnemosyne-proj-users/browse_frm/thread/433872b155ad7451/6e1ef5c773dcc5e7?lnk=gst&q=multiplication#6e1ef5c773dcc5e7
You could, of course, find some repository of bridge problems or
situations, and turn them into flashcards. A few hundred would, I'd
guess, cover all the principles of interest in sufficient volume that
you'd memorize the principles rather than the problems.
>> > (This article is about SuperMemo, but Wikipedia led me to your site as an
>> > alternative.)
This, incidentally, is why working on Wikipedia articles is useful for
FLOSS projects, even though it can be difficult. Aside from anecdotes
like this, one of the top referrers for xmonad.org is the Xmonad
article I wrote.
>> > 2) Will your product only really take off when there are large decks of
>> > cards, prepared by skilled experts, covering a range of topics? However
>> > clever and well-founded the system is, how do you reach a critical mass
>> > of
>> > content as opposed to mere method?
Maybe. Dynamic cards is my favored breakthrough, but perhaps existing
static content isn't sufficiently high quality and lacks multimedia.
Anki is a competing SRS, which is roughly twice as popular as
Mnemosyne according to the Debian Popcon:
http://qa.debian.org/popcon-graph.php?packages=anki+mnemosyne&show_installed=on
One of its signature features is integrated downloading of user decks
(rather than go to the Mnemosyne website, search around, download, and
import manually, one opens File, clicks the appropriate item, and then
can incrementally search the listings and download & install with
another click), which seems to have made contributing back one's cards
much more popular.
--
gwern
