Register  Login  Active Topics  Maps  

Readlang, my language reading site

 Language Learning Forum : Learning Techniques, Methods & Strategies Post Reply
131 messages over 17 pages: 1 2 3 4 5 6 7 ... 7 ... 16 17 Next >>
SteveRidout
Diglot
Groupie
Spain
readlang.com
Joined 1722 days ago

65 posts - 55 votes 
Speaks: English*, Spanish

 
 Message 49 of 131
03 June 2013 at 12:57pm | IP Logged 
Crush wrote:

I was thinking of something like Verbix which provides verb conjugations for a large number of
languages. Another possibility would just be to have the user manually add in the infinitive so
that the system would store them all as part of the same word (and potentially recognize the
word the next time you see it).


Crowd sourcing like this is a good idea, both for grouping word families, and improving on the
translations from google which aren't always ideal.

Crush wrote:

I would just send it as is, in the order they're in. Eg. "Nunca se puede acordar de...", i'd
just send "se acordar". For the most part google recognizes them and there's always the
possibility of checking the dictionary and modifying the definition yourself. I don't think
detecting these words should be the system's job, but it would be nice for the user to be able
to select them manually.


Still could be confusing for the UI though - the user would need to specify whether they want
the words combined or not. Isn't it OK just to translate the whole phrase including the in-
between words in this case?

Crush wrote:

In that case, i'd love to see support for Galician and Basque :)


Done! These are both online now, along with Latin which someone else asked me for :)

One note though - I don't have word frequency tables for these languages, so the most useful
words won't be prioritised for flashcard learning like they are for the more popular languages.

Crush wrote:

I realize that, i was more making a point that if there are two words you don't know, for
example, maybe i don't know what "olvidada gruta" means in the phrase "Nos adentramos en aquella
olvidada gruta..", clicking on both words will combine them into one. If you can make the
distinction between words that should be together and words that aren't, i think it'd make
adding support for "split" words/phrasal verbs a little easier.


Hmm, thinking aloud, this could be possible by sending google 3 queries:

1. "olvidada"
2. "gruta"
3. "olvidada gruta"

If the translation of (3) is the same as the translations of (1) plus (2) I could keep them
visually separate. But there'd be a tradeoff in the time and cost taken, and it gets more
complicated with longer phrases. I'm not sure it's worth it. It would more sense if in the
future I use a translation service which returns details of how each word/phrase is mapped.

(BTW - I've just updated the site to work around the 300ms tap delay on mobile devices. It makes
translating feel much quicker and nicer on my iPad, can't believe it took me so long to try
this!)



Crush
Diglot
Senior Member
ChinaRegistered users can see my Skype Name
Joined 3305 days ago

1622 posts - 682 votes 
Speaks: English*, Spanish
Studies: Mandarin, Esperanto, Basque

 
 Message 50 of 131
03 June 2013 at 1:34pm | IP Logged 
Thanks for adding Galician and Basque :)

SteveRidout wrote:
Crowd sourcing like this is a good idea, both for grouping word families, and improving on the
translations from google which aren't always ideal.
I was going to suggest something like this, though i think it might work better for verb forms and noun/adjective declensions. The system for definitions on LingQ is really cluttered, and words can have a wide variety of translations. Maybe a word could have a list of definitions, and if someone finds a definition not in there they can add a new entry to the list of definitions without making an entire new list? I'm also not sure if i want the English (translated) word to show up in the article, maybe hovering over the word to show the definition (once you've clicked on it to check the definition)? That way i can still see what the original word is that i didn't know. I'm really not sure how flexible the language it's written in is, maybe you could have the translation appear for a second and then if you want to see the translation again you can hover over the word? This probably wouldn't be as convenient for users using touch devices, though.

SteveRidout wrote:

Still could be confusing for the UI though - the user would need to specify whether they want
the words combined or not. Isn't it OK just to translate the whole phrase including the in-
between words in this case?
In the case of German, that can be a lot of words. Or English, take the phrasal verb "to take out": My mother asked me to take the trash i left downstairs when i got home from school out. You could translate the whole phrase, but that also makes the flashcards less useful, you've got to remember the entire phrase. Again, i don't know if something like this would be possible, but in line with the suggestion i had above about the 1-second delay, you could use that one second delay to let the user pick another word to attach to it. So in the previous sentence i would click on "take" then on "out". As it is, i've just been using it for Catalan and Spanish, for these languages adding the whole phrase isn't much of an issue since it's generally only an extra word or two. Though it will look up "d'aquesta", even if i just want to look up "aquesta". Editing the flashcard i can change it though, and it even updates the tests, which is nice :)

SteveRidout wrote:

Hmm, thinking aloud, this could be possible by sending google 3 queries:

1. "olvidada"
2. "gruta"
3. "olvidada gruta"

If the translation of (3) is the same as the translations of (1) plus (2) I could keep them
visually separate. But there'd be a tradeoff in the time and cost taken, and it gets more
complicated with longer phrases. I'm not sure it's worth it. It would more sense if in the
future I use a translation service which returns details of how each word/phrase is mapped.
I think it'd be nice to have a way to distinguish between individual words and group words. Either with the one second time limit or perhaps holding a special key down or right-clicking extra words. Again, i'm not really sure what's possible programming-wise nor how much work it'd be to implement them, i've just got ideas on what i'd like to see :P



SteveRidout
Diglot
Groupie
Spain
readlang.com
Joined 1722 days ago

65 posts - 55 votes 
Speaks: English*, Spanish

 
 Message 51 of 131
04 June 2013 at 9:44am | IP Logged 
Crush wrote:
Thanks for adding Galician and Basque :)

SteveRidout wrote:
Crowd sourcing like this is a good idea, both for grouping word families, and
improving on the
translations from google which aren't always ideal.
I was going to suggest something like this, though i think it might work better for verb
forms and noun/adjective declensions. The system for definitions on LingQ is really cluttered,
and words can have a wide variety of translations. Maybe a word could have a list of definitions,
and if someone finds a definition not in there they can add a new entry to the list of
definitions without making an entire new list? I'm also not sure if i want the English
(translated) word to show up in the article, maybe hovering over the word to show the definition
(once you've clicked on it to check the definition)? That way i can still see what the original
word is that i didn't know. I'm really not sure how flexible the language it's written in is,
maybe you could have the translation appear for a second and then if you want to see the
translation again you can hover over the word? This probably wouldn't be as convenient for users
using touch devices, though.


Balancing the amount of distraction with richer translation data is going to be tricky, and I may
have to add some user-specific settings eventually. Working well on touch devices is essential,
as I think tablets offer the most comfortable reading experience.

Crush wrote:

SteveRidout wrote:

Still could be confusing for the UI though - the user would need to specify whether they want
the words combined or not. Isn't it OK just to translate the whole phrase including the in-
between words in this case?
In the case of German, that can be a lot of words. Or English, take the phrasal verb "to
take out": My mother asked me to take the trash i left downstairs when i got home from school
out. You could translate the whole phrase, but that also makes the flashcards less useful, you've
got to remember the entire phrase. Again, i don't know if something like this would be possible,
but in line with the suggestion i had above about the 1-second delay, you could use that one
second delay to let the user pick another word to attach to it. So in the previous sentence i
would click on "take" then on "out". As it is, i've just been using it for Catalan and Spanish,
for these languages adding the whole phrase isn't much of an issue since it's generally only an
extra word or two. Though it will look up "d'aquesta", even if i just want to look up "aquesta".
Editing the flashcard i can change it though, and it even updates the tests, which is nice :)


This is a nice idea. But I feel to do it well with a clear UI will take a bit of time and to be
honest, I've got a bunch of other improvements things that I need to work on first so I'm not
sure when/if I'll get around to it.

Crush wrote:

SteveRidout wrote:

Hmm, thinking aloud, this could be possible by sending google 3 queries:

1. "olvidada"
2. "gruta"
3. "olvidada gruta"

If the translation of (3) is the same as the translations of (1) plus (2) I could keep them
visually separate. But there'd be a tradeoff in the time and cost taken, and it gets more
complicated with longer phrases. I'm not sure it's worth it. It would more sense if in the
future I use a translation service which returns details of how each word/phrase is mapped.
I think it'd be nice to have a way to distinguish between individual words and group
words. Either with the one second time limit or perhaps holding a special key down or right-
clicking extra words. Again, i'm not really sure what's possible programming-wise nor how much
work it'd be to implement them, i've just got ideas on what i'd like to see :P


Both a time limit and keyboard shortcuts are possible, the problem is in implementing it in a way
that most users understand - a time limit in particular could be confusing, I would probably
prefer hold down Ctr or Shift, but then would need an alternative for touch devices.

Thanks for the ideas, very interesting!



Crush
Diglot
Senior Member
ChinaRegistered users can see my Skype Name
Joined 3305 days ago

1622 posts - 682 votes 
Speaks: English*, Spanish
Studies: Mandarin, Esperanto, Basque

 
 Message 52 of 131
04 June 2013 at 2:36pm | IP Logged 
Cool, it's great to hear that those things would be possible! There's definitely no rush, as it is it's already been incredibly useful for me. I'm excited to see where it goes.

For adding in verb forms, perhaps you or someone could write a script to pull the data from Verbix (or other verb conjugation sites, there are tons of them!). It might have to be personalized for each language, though. You could then either have the user type in the infinitive to search or have volunteers preload a stock of verbs into the database. Memrise also has volunteer "curators", though their role seemed much more important in the beta site. It might be a lot of work, but volunteers for each specific language could maybe monitor both new words being entered into the database and new definitions being added to words. I'd be happy to help with any/all of these, especially for languages i'm studying, and i'm sure so would many others. It might be a ways down the line, but i'm really excited because you're working on the language-learning program i always wanted to write.

And maybe i'll need to get a touchscreen device just for this...

Also, i saw that The Little Prince for Spanish was divided into sections, is preliminary support for chapters implemented?



SteveRidout
Diglot
Groupie
Spain
readlang.com
Joined 1722 days ago

65 posts - 55 votes 
Speaks: English*, Spanish

 
 Message 53 of 131
04 June 2013 at 3:10pm | IP Logged 
Crush wrote:
Cool, it's great to hear that those things would be possible! There's
definitely no rush, as it is it's already been incredibly useful for me. I'm excited to
see
where it goes.

For adding in verb forms, perhaps you or someone could write a script to pull the data
from
Verbix (or other verb conjugation sites, there are tons of them!). It might have to be
personalized for each language, though. You could then either have the user type in the
infinitive to search or have volunteers preload a stock of verbs into the database.
Memrise
also has volunteer "curators", though their role seemed much more important in the beta
site. It might be a lot of work, but volunteers for each specific language could maybe
monitor both new words being entered into the database and new definitions being added
to
words. I'd be happy to help with any/all of these, especially for languages i'm
studying,
and i'm sure so would many others. It might be a ways down the line, but i'm really
excited
because you're working on the language-learning program i always wanted to write.


That sounds great! I've actually been tempted to start an open source project to crowd
source a database of translations, definitions, and verb conjugations like this. I
looked at Wiktionary but the pages aren't ideally structured for using programmatically
- each of the languages needs it's own parser. Designing a structure that fits all
languages may be very tough, I'd certainly need advice about this. If anyone has ideas,
or knows of an existing similar project, I'd love to hear from you!

Crush wrote:

And maybe i'll need to get a touchscreen device just for this...


I got an iPad last year and hardly a day has gone by without using it! There are more
options now though and I must say I'm tempted by the Nexus 7 since the size and weight
seem better suited for reading.

Crush wrote:

Also, i saw that The Little Prince for Spanish was divided into sections, is
preliminary
support for chapters implemented?


This is a one-off hack that I did because it looked so cluttered before. Right now I'm
working on better editing tools for people to create and share documents and this will
hopefully include chapter support.

Edited by SteveRidout on 04 June 2013 at 3:18pm



senor_smile
Diglot
Senior Member
United States
Joined 3826 days ago

110 posts - 5 votes
Speaks: English*, Spanish
Studies: Latin, Russian

 
 Message 54 of 131
04 June 2013 at 6:53pm | IP Logged 
For Latin, I would suggest using Whitaker's words dictionary located at
http://www.archives.nd.edu/cgi-bin/words.exe .

It is fantastic as it parses almost any Latin word, telling you its root form, possible
declensions and definitions.



Crush
Diglot
Senior Member
ChinaRegistered users can see my Skype Name
Joined 3305 days ago

1622 posts - 682 votes 
Speaks: English*, Spanish
Studies: Mandarin, Esperanto, Basque

 
 Message 55 of 131
05 June 2013 at 6:39am | IP Logged 
Another suggestion: it might be useful to have a page with a list of the languages you've currently started studying and how many cards are do. Then you can click on that language and it'll take you to the reviews page. If you've got several different languages it might start to be a little difficult to keep track of them all.

For pulling data from online sites, it might be better in the end to do it on an individual language basis. You (or volunteers) could enter a list of all the names of the tenses and use that list to pull the data out. Since it's different for every language it'd have to be personalized, i think. The database that stores the information at least would be different for each language. Verbix, for example, has each tense separated within a <td> tag. You could search for the tense name (which will be bolded) and then read the rest of the text in the <td> tag. Most of the languages i've looked at also include the pronouns next to the word so you could probably strip those.



SteveRidout
Diglot
Groupie
Spain
readlang.com
Joined 1722 days ago

65 posts - 55 votes 
Speaks: English*, Spanish

 
 Message 56 of 131
05 June 2013 at 11:12am | IP Logged 
Crush wrote:
Another suggestion: it might be useful to have a page with a list of the languages
you've currently started studying and how many cards are do. Then you can click on that language
and it'll take you to the reviews page. If you've got several different languages it might start
to be a little difficult to keep track of them all.


Great idea! I'll work on this.

Crush wrote:

For pulling data from online sites, it might be better in the end to do it on an individual
language basis. You (or volunteers) could enter a list of all the names of the tenses and use
that list to pull the data out. Since it's different for every language it'd have to be
personalized, i think. The database that stores the information at least would be different for
each language. Verbix, for example, has each tense separated within a <td> tag. You could search
for the tense name (which will be bolded) and then read the rest of the text in the <td> tag.
Most of the languages i've looked at also include the pronouns next to the word so you could
probably strip those.


I need to do some research on this before discussing the best approach. I'd like to standardise
the format as much as possible across languages, so tools like Readlang don't have to implement
language specific code. Give me a bit of time and I'll start a new thread to discuss this.

senor_smile wrote:
For Latin, I would suggest using Whitaker's words dictionary located at
http://www.archives.nd.edu/cgi-bin/words.exe .


Thanks, looks useful.



This discussion contains 131 messages over 17 pages: << Prev 1 2 3 4 5 68 9 10 11 12 13 14 15 16 17  Next >>


Post ReplyPost New Topic Printable version Printable version

You cannot post new topics in this forum - You cannot reply to topics in this forum - You cannot delete your posts in this forum
You cannot edit your posts in this forum - You cannot create polls in this forum - You cannot vote in polls in this forum


This page was generated in 0.5308 seconds.


DHTML Menu By Milonic JavaScript
Copyright 2017 FX Micheloud - All rights reserved
No part of this website may be copied by any means without my written authorization.