[Parley-devel] Fwd: Features in next version of Parley

[Inge Wallin]

On Thursday, September 11, 2014 09:03:01 AM Anša Vernerová wrote:
> Hi,
> 
> >> Can you build Parley from the sources?
> >> 
> >> I will give it a try (during the weekend).
> > 
> > How did it work out for you? Where you able to get it running?
> 
> Yes, I have it working pretty nicely. I must admit that in recent
> weeks, I have spent more time writing down my wishes and ideas than
> actually practicing vocabulary. Should I concentrate on testing some
> specific feature? Or track my progress in some way?

haha, haven't we all been there?  I, myself, is back in Europe now but I had a 
bad disk krasch so I had to borrow another computer until my own is fixed. To 
be honest we haven't implemented any new feature lately.

But there is one thing I would like your opinion of, namely Amarvir's GSoC 
project.  It's in the lessoncreator branch. I don't consider it merge ready 
yet, some more work is needed. But this could be a perfect way for you to give 
some input before things are finalized.

To check out a branch type:

  git checkout lessoncreator

And then build just like normal.  To go back to the master branch, do:

  git checkout master

and rebuild again. You may have to add some dependencies to make it build, I 
am not sure right now. But if so we will help you of course.

> >> In the following scenarios, in case of a failure to produce the right
> >> answer, the learner has to indicate which confidence level should be
> >> affected:
> >> L2.text/L2.sound -> L3.text/L3.sound     (the problem may be both on
> >> the L2 and on the L3 side, user clarification is necessary, then the
> >> confidence levels for the passive skill of L2 and the active skill of
> >> L3 can be updated)
> > 
> > This is where it gets interesting.  Is the L2 -> L3 scenario really
> > worthwhile? All theories about learning languages that I have read
> > indicates strongly that you need context to learn something well. And to
> > have 2 words that supposedly mean the same thing in 2 languages that
> > you don't know sounds to me like the oppositve of having context.
> 
> I was mostly thinking of these scenarios:
> 
> 1) My files contain three columns:
> Target language word or phrase - Mother tongue equivalent - Target
> language explanation or definition
>       I only fill the last column if I can come up with (or, more
> often, find in a learner's dictionary) a synonym or a definition that
> makes at least some sense to me. This way, I am practising the target
> language in the context of the target language. However, it sometimes
> happens that I forget the meaning of the synonym or I confuse it with
> a similar sounding word - a mistake which I recognize as soon as I see
> the correct answer. (I use the third column only for questions; I do
> not practice the opposite direction.)
>     Conceptually, I suppose, my third column should go into the
> "Synonym" or "Paraphrase" fields. Besides there being no way of
> practicing "paraphrase -> actual words", I still find it easier to use
> the paraphrases as if they were L3.
> 
> A similar case would be with question-answer pairs in the Target
> language, where the Question would mostly be easily understandable but
> could sometimes contain problematic vocabulary, and the answer would
> require some factual knowledge. ("When is Christmas celebrated in the
> United Kingdom?" - "On December 25." The user may confuse Christmas
> with Easter or the United Kingdom with the United States, as well as
> not remember the answer.)

yeah, this looks useful

> 2) (I intend to do this, but I have not tried it yet) I was thinking
> of trying to memorizing Biblical verses in my mother tongue, using
> Latin as the source language. I've had some introduction to Latin and
> after all, quite a few Latin roots appear in vocabulary of other
> languages I have been studying, so although my Latin is not good
> enough for me to actually read in it, I can take guesses, and after
> seeing the actual Czech translation, I can tell which word in Latin
> corresponds to which word in Czech. (Most of the time, that is.) So
> the idea is to get exposed to Latin while memorizing sentences in my
> mother tongue. Paradoxically, the context is provided by the language
> that would probably be marked as "Learning" in the current Parley.

Hmmm... ;)

> 3) I am a rather advanced learner of German, so using German instead
> of my mother tongue would make good sense: German items would get
> reinforced in my brain while I would be learning another language.
> However, my passive knowledge of German is much better than my active
> knowledge. Thus, in testing Learning->German in a written practice, it
> can easily happen that I will know the meaning of the word (and be
> able to confirm that after seeing the German equivalent as the
> answer), but actually not be able to write down the German equivalent
> before seeing it. In this case, the language marked as "Known" is
> known only passively and it makes sense to record its confidence
> levels for the active skills of writing and speaking.

Yeah, this is how I use English.  I can hardly find any lesson material with 
Swedish so I use English as my "native" language. Except I normally also know 
the word actively.

> >> At an initial stage of learning, the learner might also want to
> >> practice just the following two scenarios. Should they have their own
> >> grades?
> >> dictation:            L2.sound -> L2.text
> >> 
> >>       the learner knows how to capture the sound in writing, but
> >> 
> >> possibly does not know what it means
> >> pronunciation:     L2.text -> L2.sound
> >> 
> >>       the learner remembers how to pronounce the word, but possibly
> >> 
> >> not its meaning
> > 
> > Technically I think they should. We should not design away any
> > possibilities in the format alone. But in practice I also doubt that this
> > is a good way to train and I also question if it's actually possible.
> 
> I was mostly thinking of people who are learning a language that uses
> a different alphabet (in the sense of a relatively small set of
> characters that directly correspond to the sounds of the target
> language). I am currently learning Hebrew, which is an example of such
> a language. Most Hebrew courses start with a chapter on reading,
> although the student does not actually understand any of what they are
> reading - the idea is to spend some time learning the letters and
> sounds.

Ok.  As I said I think it should be technically possible and after that you 
can use it as you like.

> > What is actually happening is that if you learn a new word from scratch it
> > goes from grade 0, pregrade 0 to grade 0, pregrade 1. And then when you
> > train it repeatedly with intervals 3.5 minutes, 7 minutes, 15 minutes, and
> > so on, it goes from pregrade 1 to pregrade 7.  The next step is grade 1,
> > pregrade 0 and from there it's exactly as before, i.e. pregrades are not
> > used anymore.
> 
> What happens when I only practice once a day? It seems to me that in
> the case that I get the word right every single time (with the
> exception of the very first one), I get to see it on seven consecutive
> days.

It depends on how long you practice each time. The selection is done at the 
creation of a new session.  If you (say) practice one hour per time then you 
will run 10s of sessions. Then it's not at all unlikely that you will get the 
same new word in more than one session, especially in the beginning. So this 
means that you will see the word fewer than 7 days in a row before coming up 
to the long-term stage. If you practice for a very short term or have very 
many words to practice things may be different.

What happens when you practice so little time that you don't have time to 
exhaust all the words is a question of strategy. Right now the strategy is to 
choose words with higher grades first, protecting your investment in training 
these words, so to speak. But we have discussed other strategies too.

> I suppose the new training system proposed by Andreas will
> somehow address this, but for the time being, I think that the word
> should skip all levels that have shorter intervals than the one for
> which the word was actually retained. (This does not apply only to
> pregrades.)

I am not sure I understand. Do you mean that if you have a word at for 
instance grade 4, fail to recognize it and it falls down to grade 0, pregrade 
1 then when you recognize it again it should get back up to grade 4?  That is 
a very interesting idea. I haven't thought about that because I don't mind 
training some words a little extra.

> Also, it seems to me that once I answer the word correctly, it will
> not appear in the same session again, even if the interval for next
> repetition comes about. (I usually have sessions lasting longer than
> 15 minutes.) A new session has to be started.

This is correct. How many words do you typically have in a session?  It could 
be a good enhancement to reintroduce words into the session when they "time 
out" so to speak even when the session is not restarted.

> Moreover, as the "max
> number of new words" cannot be set below 2, 2 new words will be
> introduced if a new session is started right after a previous one.
> (Only if there are some new words left in the lessons
> marked for possible practice.) I would welcome some kind of "go back
> to practice words from this session that are scheduled for now" button
> that would appear at the end of any session. Or maybe "practice all
> initial stage words that are scheduled for now". Aha! This latter
> case, probably more reasonable, can be achieved with the current
> practice setup (there is a setting for maximum confidence level). It
> might be a bit cumbersome, but it would work. However, setting both
> the minimum and maximum confidence levels to 0 still does not
> distinguised between new words and initial stage words.

Yep, all this is true.  We have discussed to introduce a limit on the total 
words in the initial stage (grade=0, pregrade>0) so that no new words are 
introduced once this limit is reached until at least one word graduates to the 
long-term stage with grade >=1.

But I guess that if you run long sessions then these two ideas (limit and go 
back to earlier words in the same session) don't exclude each other.

> > The idea is that the sentences will give you the structure of the language
> > rather than individual words. You will learn the vocabulary anyway since
> > many of the sentences use the same word in different contexts.
> > 
> > It's an interesting idea and I am looking into it if we can incorporate it
> > into Parley at some point in the future.
> 
> I always copy the whole text of the lessons from
> textbooks/courses/phrasebooks into parley, separate it into items and use
> it for practice. If it is possible, I use the sound recordings that come
> directly with the
> original materials - I am an aural learner.
> 
> I am not sure if Parley could really do more for me in this respect,
> because automatically generated sentences would have to be equipped
> with automatically generated sound, and it is unlikely that this
> automatically generated sound would be satisfactory. In English, for
> example, word stress changes depending on the sentence context. This
> is important, because learners actually remember the "melody" of words
> and phrases before they fully learn them. (I read just today that it
> was shown already in the 1960's or 1970's that learners could
> correctly answer questions about the number of syllables and accent
> patterns of vocabulary items, even though they were not able to recall
> them correctly.)

Something that I have just become aware of is Glossika's mass sentence 
concept.  I am very intrigued by it (www.glossika.com). You have to pass the 
front page since that is very sales centered. But the concept is very 
interesting and we might do something like that in Parley.

> My feeling is that users should be encouraged to fill their
> collections with sentences that occur in whatever materials they are
> working with. They also should be encouraged to use definitions and
> synonyms in the target language instead of translation equivalents in
> their mother tongue (unless they intend to be translating between the
> two languages a lot, in which case it makes more sense to strengthen
> the associations between the two languages). All this provides natural
> structure and context.

After looking into the work of Glossika, I tend to agree 100%.

> A different topic: my current reading has stressed the importance of
> mnemonic techniques for the early stages of learning a vocabulary
> item. I think maybe I could write some kind of summary of that chapter
> for the parley documentation, if that would be useful. It occurred to
> me that you could support users in making up their own mnemonics by
> renaming the "Comments" field into "Mnemonics and comments".

I don't know.  I think the current quick repetition pattern with the pregrades 
(we really shouldn't use that name outside the developer discussions) make 
learning extremely efficient. I can learn the vocabulary that I study very very 
fast compared to other methods that I have tried.

That said, we are very grateful for help with the documentation, of course. 
Another thing I was thinking of is to have a part of parley that explains the 
concepts and methods behind the implementation so that the user can get it 
inside Parley and don't have to go read a manual. I'd also like to pepper the 
UI with lots of 'what's this'-help which KDE has good support for.

	-Inge

> Anša
> _______________________________________________
> Parley-devel mailing list
> Parley-devel at kde.org
> https://mail.kde.org/mailman/listinfo/parley-devel