In this following paragraph (1), I wrap up my discussion of a problem fundamental to the flashcard-program world that is usually ignored, and I prescribe a solution.
1. I want to clarify that when I said "I see wǒ 我 and rén 人 a hundred times every day" I was referring also to my studies outside of mnemosyne. Much of my last few posts is partly what I've come up with after thinking about how this outside exposure should be or could be accounted for in a flashcard program. Briefly: On the one hand, cards are promoted without the justification the algorithm assumes. (Seeing 我 every day undercuts the _meaning_ of the interval and the increment on a card that the algorithm thinks I haven't seen for 120 days! It has made a false assumption!) But on the other hand, not promoting can't be right either. My solution lies in ensuring that 我 is represented on more-composite cards that are still being mastered, and that are less likely to be seen literally every day, while the simple card for 我 itself becomes a "don't care". Here is a perfect illustration: I don't think I've opened volume 3 in my textbook series in weeks. Now, if volume 3 is represented by nothing more than its atoms, like 我, then all I have, within this flashcard system, to show that I've mastered volume 3, are the intervals on these cards, intervals that I argued above are unjustified, with respect to volume 3, *because I've probably been seeing these items daily in all subsequent volumes*. But, if the card for 我 is pruned or is a "don't care", while 我 is still represented on a card that shows, let's say, the entire lesson in which 我 first appears, then mnemosyne *opens volume 3 for me, to a lesson I haven't seen in weeks, on the schedule I command*, I prove that I still know that whole lesson, and 我 is validly along for the ride: my zealously guarded hoard of characters is "all present and accounted for". 2. If there is a linkage such that promoting for example this card, "我不 是中国人.", also auto-promotes the simple card for 我, then yes as you said "this is almost equivalent to automatically pruning them". Yes, it would be unnecessary to actually remove it... very convenient. 3. > However the idea of zooming out to higher knowledge abstraction levels > as the 'atomic' cards become well-learned is very interesting, > although it certainly seems like a difficult and topic-specific task > that is beyond the scope of current SRS systems which require little > information from the user and usually ignore the card contents > completely. The *automatic* parsing of a text? Yes, it would be topic specific. I was exploring the extreme... But it wouldn't be necessary that the flashcard program know the subject on the cards; assuming that the subject can benefit from a hierarchical breakdown (most subjects) what matters is only the relationship between the cards, which would be reflected in their locations in a tree. I *think* the program author said that v2 will have hierarchical categories. I *think* this would serve as the tree. [With one caveat: I think? it isn't technically exactly a tree if the leaf nodes point back to more than one parent. Here's what I mean: under a node labeled say "volume 3" (broken down optionally into) "lesson 2" (broken down optionally into) line 5, might be "我不是中国人." with one leaf being 我. While another line in another volume will also have 我 as a leaf. Therefore, 我 is more than a simple leaf, though I forget the terminology. It would be necessary that once 我 is declared pruned (or is being auto-incremented) in one place, it should prune or auto-increment itself everywhere, indicated maybe by changing color in the category browser. I'm not an expert, but I believe the software for managing relationships like this is routine, so the code is probably in a library somewhere. At a higher level, this is just database management, right?] In conclusion, let me bring this back to a practical level. Most of my last few posts are arguing about issues that fall somewhere between a flashcard program, and a lesson plan or study strategy (divide and conquer while naturally observing dependencies...doesn't this happen in just about every sphere?). The most important part was my argument for increasing-complexity as warranted (such as it is for me in Chinese, now). So far, this doesn't necessarily involve the flashcard program, but I offered some non-expert ideas on how it could be supported, if anyone accepts the principle. The best news is that v2 (assuming I'm remembering right that it will permit hierarchical categories) will already have a structure that will support managing these matters *manually*, and incidentally scrolling windows too. Perfect! I'm doing it manually now, following my common sense with regard to how to study big subjects where simplicity is folded into complexity. It won't surprise you that my original interest in trees followed from my obsession with chess and then chess programming. Math, chess, language, same thing: "One (tree) to rule them all". On Jul 30, 7:44 pm, Oisin Mac Fhearai <[email protected]> wrote: > On 30 Jul 2009, at 22:57, querido <[email protected]> wrote: > > > > > I see wǒ 我 and rén 人 a hundred times every day! I feel safe assumi > > ng > > that as long as hundreds of sentence-cards (with audio by the way) are > > present which contain these characters, I may prune them. (The Author, > > above, used this word; I wasn't even considering it.) The task then is > > to develop some reliable rule to govern this. Then, optionally > > *automating* the action of the rule led me to the ideas I wanted to > > post. > > In these cases, you would rate the card with the maximum score > repeatedly such that the intervals increase rapidly - this is almost > equivalent to automatically pruning them (e.g. In my deck, words like > 人 and 是 were assigned intervals of a few years very quickly - they > are, for all practical purposes, pruned, by virtue of the rapidly > increasing long intervals and short answer time). > > However the idea of zooming out to higher knowledge abstraction levels > as the 'atomic' cards become well-learned is very interesting, > although it certainly seems like a difficult and topic-specific task > that is beyond the scope of current SRS systems which require little > information from the user and usually ignore the card contents > completely. > > Oisín > > > > > On Jul 25, 8:46 pm, Ben <[email protected]> wrote: > >> Thanks, this is an interesting email. I agree with your thoughts > >> that I quoted below. I think a program like Mnemosyne seems pretty > >> optimal if your goal is, for the rest of your life, to be able to > >> recall a fact at random in a few seconds from your deck. For > >> languages, this may fit real life very well---you never know when > >> you will run across some word and you want to be to recall the > >> meaning in a few seconds. > > >> However, this context-free benchmark may not appropriate for some > >> other knowledge. For instance, suppose I wanted to remember linear/ > >> abstract algebra for the rest of my life well enough that, should I > >> run across a paper that uses basic linear algebra, I could spend 10 > >> minutes reviewing and be able to understand the paper. This is > >> different from being able to remember in a few seconds the > >> definition of a "homology" or whatever. > > >> This isn't a complaint about Mnemosyne of course, I'm just agreeing > >> that a card-based system may not be the ultimate answer to > >> retaining all knowledge and skills. > > >> -- > >> Ben > > >> ----------------- Original message ----------------- > >> From: querido <[email protected]> > >> To: mnemosyne-proj-users <[email protected]> > >> Date: Fri, 24 Jul 2009 16:40:07 -0700 (PDT) > > >> ... > >> 1. One resists the idea of pruning cards, but after thinking about it > >> I have found a good rationale (not proof). Some have opined that at > >> very short intervals, beginning from the first glance at a fact, > >> flashcards are not yet the ideal tool, that one should first learn > >> the > >> fact to some (arguable) degree. (I agree, without knowing how well > >> that should be. I've tried stretching this out as long as eight days, > >> memorizing material before flashcarding it for retention-only. An > >> ideal is probably in there somewhere.) Now, similarly, maybe the > >> single-fact, atomized-data style flashcard system becomes non-ideal > >> again at long intervals too. For example, one of my earlier Chinese > >> textbooks broke down into 900 flashcards total, which are still in > >> mnemosyne. I can now read, aloud or not, this book fairly rapidly, > >> and > >> understand its audio. So, reading or listening I zoom over hundreds > >> of > >> "atoms", all nicely connected with context and grammar, etc.- real > >> language. At some point, it might be a good idea to prune all 900 > >> cards and make a "review scheduling style" card maybe like this: > >> Front- "read Modern Chinese Reader aloud" Back- "Did you know > >> (almost) > >> everything?" (Where "almost everything" concedes that your brain is > >> not a machine, after all; we all have a standard, and compromise on > >> "perfection" for the sake of just carrying on living and learning.) I > >> now intend to do this when I get around to it. (By the way, this > >> would > >> make it even more important that you're learning from something > >> cohesive, like a book with lots of context, *so that* you could later > >> prune all of the cards, knowing you can still hold them all securely > >> in one hand.) > >> I had thought that once cards were known perfectly well that each > >> card > >> would become sufficiently effortless. It doesn't; it is still many > >> times harder than flying over that same fact in context while reading > >> or listening. (You could prove that.) I had also thought that once > >> they were promoted far enough, they would practically disappear. > >> Well, > >> your testimony above confirms my impression that they don't, quite > >> well enough. This is what motivated me to think about this again. > >> ... > > --~--~---------~--~----~------------~-------~--~----~ You received this message because you are subscribed to the Google Groups "mnemosyne-proj-users" group. To post to this group, send email to [email protected] To unsubscribe from this group, send email to [email protected] For more options, visit this group at http://groups.google.com/group/mnemosyne-proj-users?hl=en -~----------~----~----~----~------~----~------~--~---
