Lingocracy Learn by reading what you like (Learning Techniques, Methods & Strategies) Language Learning Forum

Lingocracy Learn by reading what you like
Tags: Reading
Share with: Delicious Digg reddit Facebook StumbleUpon
Language Learning Forum : Learning Techniques, Methods & Strategies

134 messages over 17 pages: 1 2 3 4 5 6 7 ... 4 ... 16 17 Next >>

Crush
Tetraglot
Senior Member
ChinaRegistered users can see my Skype Name
Joined 5667 days ago
1622 posts - 2299 votes

Speaks: English*, Spanish, Mandarin, Esperanto
Studies: Basque

Message 25 of 134

15 February 2014 at 9:49pm | IP Logged

I'm excited to hear that #1 can be implemented. It would've helped me a lot in Russian when trying to figure out cases and aspect :)

I also realize #2 is next to impossible to accurately detect, that's why i thought it could be left up to the user to find. For example, i could click on A then B and have it be treated as one entry, and perhaps it would show up that way for other users who read the same article. You could even have the user type in how it should be represented if it's still hard to analyze, as for some languages the word would be AB and for others BA. I also am not sure how "language-specific" you're trying to or not to be. For example, it appears that Thai requires a bit of extra work to make it compatible, and it looks like Mandarin and possibly Japanese have had some special work done to them to get them running. Maintaining 50 languages is no small feat, especially when they have language-specific code. But i would be more than happy to do some work to get at least the languages i'm studying up and running (collecting word/frequency lists, conjugation/declension charts, etc.) and afterwards help out with whatever else still needs some work.

You could use the Wordreference dictionaries, they're available for quite a few languages. I can also share some really nice online dictionaries for Basque and Catalan that can be queried by adding the word in the URL. You'd have to process them first, though, as they don't handle conjugations or cases. I believe the Wordreference dictionaries do, at least for Spanish and French (and possibly Italian).

Anyway, like i said, i'd be happy to put some stuff together to help the project advance.

EDIT: It's also a bit hard to select multiple words. The only way i can find to do it is to click on a word then highlight other words. Could you perhaps have it work the other way, too? Selecting some words then clicking on them? Though that might not work since you the highlighting gets erased when a word pops down. I don't know what's possible, but perhaps holding [Ctrl] and then clicking on another word would add that word to the main one. Holding [Shift] would select all words between the main word and the word you clicked on. I'm also writing after using it mostly for Chinese, so my thoughts might not work for other languages that well.

A quicker way to separate joined words would be nice, too. Perhaps a little [x] or something could pop up when you hover a word group?

It's also a bit hard to select the "Split compound word" option. Sometimes it disappears when i move the mouse down before it reaches the box.

EDIT2: I also can't seem to join two words that are separated by a line break.

Edited by Crush on 16 February 2014 at 3:58am
2 persons have voted this message useful

Henkkles
Triglot
Senior Member
Finland
Joined 4055 days ago
544 posts - 1141 votes

Speaks: Finnish*, English, Swedish
Studies: Russian

Message 26 of 134

16 February 2014 at 6:21pm | IP Logged

Were it possible to tag some languages to a sidebar thingie so that I could switch (more) easily between my active languages instead of having to go through the list of languages each time? Like pinned languages, or just have all the languages in the sidebar, as it stands, around half of my full HD monitor's surface area is empty, everything clumped to the middle. I would love if I could have the hovering definition thing in the right side of the text, as I don't like the fact that it's hovering on top of the text.

I don't know what sort of changes you have planned for your interface but I hope these issues will be addressed somehow.

Also, do you intend to monetize the service at some point?
3 persons have voted this message useful

Yaan
Triglot
Groupie
France
Joined 3876 days ago
61 posts - 88 votes

Speaks: French*, English, Mandarin
Studies: Spanish, Esperanto

Message 27 of 134

17 February 2014 at 12:25am | IP Logged

I'm amazed again by those precious feedback, thanks!!

Just to give some context, "Grouping nearby words" is how word grouping word currently. Example, you can group "United" and "States" to form "United States".

Going back to our second case, let's call it "Grouping distant words" (#2), so it's easier for everyone to follow.
I'm brainstorming around this same as you, and I just see another obstacle to this: false positive. Example:
1) "I call your yellow dog out". (phrasal verb "call out")
2) "I call your dog to go out."
In 1), we should group "call" and "out" because it's the phrasal verb "call out". But in 2), that's not the case. There is no easy way to tell when it should be grouped.

False positive exist for "Grouping nearby words" too: "We have -> united states <- one by one." "united states" should not be grouped here. But this shouldn't happen often for most
languages.

--
Yes, "language-specific features" are hard to deal with. But the website is called Lingocracy (Language + Democracy), I guess that we don't have choice but being open in terme of language
support :)
Thank you Crush for your proposition. You are already helping a lot with this discussion.
At the current state, what we need most is dictionary entries for some languages. But, we can't use any online dictionaries, since they usually have some restrictive licences. Free dictionaries
like CEDICT and EDICT are just wonderful, I wish there is one like that for every language, but that's not the case.
On the interface side, we may create a simple interface that will allow benevolent people like you to add a list of dictionary entries easily.

Crush wrote:

Anyway, like i said, i'd be happy to put some stuff together to help the project advance.

Thank you!!!

Crush wrote:

It's also a bit hard to select multiple words.

Currently, it always work when you start selecting from the middle of the first word and finish the selecting at the end of the last word.
But if you start at the edge of the first word, sometime it doesn't work, we have to fix it ASAP.

Crush wrote:

I also can't seem to join two words that are separated by a line break.

This is on purpose. Is there any language where this is required? :)

Crush wrote:

A quicker way to separate joined words would be nice, too. Perhaps a little [x] or something could pop up when you hover a word group?

Good idea, we will experiment that. I'm just affraid that this will create some sort of visual clutter, even if it's on hover.
Are you splitting words very often? In my own use case, I don't split words very often.

--

Henkkles wrote:

Were it possible to tag some languages to a sidebar thingie so that I could switch (more) easily between my active languages instead of having to go through the list of
languages each time? Like pinned languages, or just have all the languages in the sidebar, as it stands, around half of my full HD monitor's surface area is empty, everything clumped to the
middle. I would love if I could have the hovering definition thing in the right side of the text, as I don't like the fact that it's hovering on top of the text.

I don't know what sort of changes you have planned for your interface but I hope these issues will be addressed somehow.

Also, do you intend to monetize the service at some point?

Yes, Crush and Serpent suggested something similar previously. Making it easier to switch language is difinitely planned. I also learn more than one language and currently it's at bit tedious to
switch, especially for the "unfeatured" languages.

Thank you for the suggestion. I guess, we should let users choose where they want the popup to appear. We will work on that.

Lingocracy will stay free similarly to Evernote or Dropbox. That means, a fully featured free version, like those two examples, with some premium features. Democratizing language learning is in our
ADN, so, the philosophy is to keep Lingocracy as accessible as possible while making the webiste sustainable. I will give more details later, but I think you got my point. Thank you for asking
the question, I didn't mention that point ealier because it's work in progress :)

--

Crush wrote:

EDIT: And how about the ability to edit/see a list of texts/URLs you've added? And perhaps the option to give a manual title (and translation) of the text? I like that it's really
simple, but it's not always that accurate. Another option would be for users to suggest translations to the title. You could either manually check them, have volunteers to check them, have a
voting system (though i don't think that's always that accurate), or have the user who shared the text be notified of the suggestion and optionally choose to approve or deny it, though users
don't always stick around that long.

Now you can edit article when adding them (title, content, and more). Of course, the goal is not to alter severly the content, but just correct some formatting issues.
There is no moderation system yet, we may implement one later, but for the moment, let's keep it simple and trust users. We still have a repport button at the of each article if needed.

Edited by Yaan on 17 February 2014 at 12:31am
1 person has voted this message useful

Crush
Tetraglot
Senior Member
ChinaRegistered users can see my Skype Name
Joined 5667 days ago
1622 posts - 2299 votes

Speaks: English*, Spanish, Mandarin, Esperanto
Studies: Basque

Message 28 of 134

17 February 2014 at 3:02am | IP Logged

Yaan wrote:

Crush wrote:

I also can't seem to join two words that are separated by a line break.

This is on purpose. Is there any language where this is required? :)

This happens to me all the time in my Chinese texts. Maybe it's due to the way it's formatted when the article is processed? See this text, for example: 撒哈拉的故事 01. There were quite a few phrases that were cut off in the middle.

As for separating words, i'm not doing it really frequently, but there have been a few times where a character was attached to the previous one when in that context it should've been connected to the next one. Or at least make it a one-click thing in the main dictionary entry, rather than having to go to "more".

Yaan wrote:

That's why my idea is that when you open up a text, the site doesn't try to process whether or not two separated words are related. That's left up to the users. If a user has marked two words as being related, perhaps THEN the site would treat them as connected, but only for that article and that instance of the word. That way, other users see that it should be connected but the site doesn't have to worry about processing them. The biggest issue with this i see is how to process it once you know they're connected. For example, in English it would be "carry out", whereas in German they would be joined as one word, with the second word preceding the first: "komme mit", would become "mitkommen". That's one thing that would likely have to vary from language to language, though you could just have the user type in what the dictionary form should be.

And i have a question about the dictionary entries. I've added a few entries to the Mandarin dictionary, however i've been getting my definitions from other online dictionaries (or from my Pleco dictionaries). I'm not sure how licenses and things work and if it's "ok" to pull dictionary entries this way. I wouldn't mind working on some dictionaries for Basque and Catalan, but i'd just be adding entries from other dictionaries which i'm not sure if that defeats the whole purpose. I haven't tried it yet, but is there a way to modify an existing dictionary entry? I'm not quite sure how entries should be organized, whether one entry per translation or one all-encompassing entry. I've been doing the all-encompassing type, adding in several translations separated by a /.

For Catalan, there's http://www.catalandictionary.org/ which is licensed under a Creative Commons License. ("© 2003-2012 This work is licensed under a Creative Commons Attribution-Share Alike 2.5 License.")

EDIT: Also, what about a way to edit texts we've already added? A lot of my texts have odd line breaks in them, i suppose that's why i was having trouble with words spanning lines. Perhaps a way to share the texts in "My Texts" would be nice, too. To me, it seems to make sense that EVERY text i add should go into my texts, perhaps you could have "My Texts" with a "Private Texts" subsection? I don't know if it was there before or not, but i do like being able to select "my texts" in the reading section :)

Edited by Crush on 17 February 2014 at 3:13am
2 persons have voted this message useful

Expugnator
Hexaglot
Senior Member
Brazil
Joined 4968 days ago
3335 posts - 4349 votes

Speaks: Portuguese*, Norwegian, French, English, Italian, Papiamento
Studies: Mandarin, Georgian, Russian

Message 29 of 134

17 February 2014 at 9:22pm | IP Logged

Add Georgian to the list of languages and we can talk :)
Maybe you could add Papiamento too, that'd be easier for me to add my own text.

Edited by Expugnator on 17 February 2014 at 9:25pm
3 persons have voted this message useful

Crush
Tetraglot
Senior Member
ChinaRegistered users can see my Skype Name
Joined 5667 days ago
1622 posts - 2299 votes

Speaks: English*, Spanish, Mandarin, Esperanto
Studies: Basque

Message 30 of 134

18 February 2014 at 5:04am | IP Logged

First off, i just noticed that you can now switch between your languages. Cool! However, on the page to change languages, the text seems to be cut off:

Quote:

We strongly recommend using English as dictionary (more entries and better quality). Of course, if

Also, why does Catalan use the Canadian flag? And when i want to change languages, there's an X that appears. I assume that's to remove the language from the list, which is a nice feature too. However, for me (Firefox) it shows up beneath the language and is impossible to click since as soon as i go down to click on it it disappears when the mouse leaves the highlighted option.

Now that it's easier to change languages, i'd like to figure out how we can start implementing some more dictionaries for the languages i'm studying :)

The new process for adding texts is great, too!
2 persons have voted this message useful

Yaan
Triglot
Groupie
France
Joined 3876 days ago
61 posts - 88 votes

Speaks: French*, English, Mandarin
Studies: Spanish, Esperanto

Message 31 of 134

18 February 2014 at 6:33pm | IP Logged

Warning, long answer:

Expugnator wrote:

Add Georgian to the list of languages and we can talk :)
Maybe you could add Papiamento too, that'd be easier for me to add my own text.

Papiamento added, have a look ;) Of course, there are no content yet, but it will be very helpful if you can add some.
As for Georgian, I need some help though. Could you help by telling me from this list which caracters should be considered as "letters" that compose words? If possible with ranges like this: 10A0-10A3, 10A5-10A8 (hexadecimal)
http://jrgraphix.net/r/Unicode/10A0-10FF

Crush wrote:

Yaan wrote:

Crush wrote:

I also can't seem to join two words that are separated by a line break.

This is on purpose. Is there any language where this is required? :)

This is because the original article uses the <pre> tag (pre tag keep the plain text formatting: new lines, spaces, etc). Which is not a very good practice for formatting this kind of text, but that's another story.
We tweaked a little bit our parsing code to deal with this case. Now, line breaks are removed for languages without space (Chinese, Japananese) and replaced by a space for other languages. (For language with spaces, we can't just remove line
breaks, if we do that: "good[line break]morning" will become a single word "goodmorning". Which is not desired.)
So now, it should work well with the Chinese example text you mentioned above.

Crush wrote:

As for separating words, i'm not doing it really frequently, but there have been a few times where a character was attached to the previous one when in that context it should've been connected to the next one. Or at least make
it a one-click thing in the main dictionary entry, rather than having to go to "more".

Got it. We will make the process simpler.

Crush wrote:

I guess you are talking about "Grouping distant words" here.
If implemented on word instance basis, well, we have to deal with other kind of issues:
1. it will require a substantial amount of work to group the words since this is must be done per article, per word instance.
2. lack of user input. Because of 1. the number of people grouping a specific word instance will be quite rare, and it's hard to validate (e.g. 10 users have marked this two words as linked, so we can make it default.)
3. making the application work on per word instance basis will require a lot of work, if feasible.

Here is another way to deal with "Grouping distant words", we can implement this at in a global way (not dependent to a text).
1. We know that "call ... out" should be grouped (with help of users or using some sort of premade list).
2. At every instance of "call" followed by a nearby "out", when hovering "call" (or "out") you can view the dictionary entries of "call" as usual, but you can also choose (button somewhere) to view "call ... out" dictionary entry (Even in case of false
positive, since we don't know).
It's up to the learner to deal with false positive, but I think this is not a huge problem, a) because of the low frequency, b) everyone is already doing that all the time, when we click the word "Apple", we see two dictionary entries, one about the
fruit, another about the Apple company, it's up to us to choose.
The benefits of implementing it at global basis (vs per text or word instance) is that a group of words need to be marked/validated once and it will work on every text and benefits to everyone.

Crush wrote:

And i have a question about the dictionary entries. I've added a few entries to the Mandarin dictionary, however i've been getting my definitions from other online dictionaries (or from my Pleco dictionaries). I'm not sure how
licenses and things work and if it's "ok" to pull dictionary entries this way. I wouldn't mind working on some dictionaries for Basque and Catalan, but i'd just be adding entries from other dictionaries which i'm not sure if that defeats the whole
purpose. I haven't tried it yet, but is there a way to modify an existing dictionary entry? I'm not quite sure how entries should be organized, whether one entry per translation or one all-encompassing entry. I've been doing the all-
encompassing type, adding in several translations separated by a /.

Yes, dictionary entries with restrictive licenses should not be added.

We think that the kind of entries used in the CEDICT are the best.
中国 -> [Zhong1 guo2] China / Middle Kingdom
Here the dictionary entry is China and Middle Kingdom. Basically, it's just a list of possible translations. I guess anyone can add this kind of dictionary entries without the need to "copy" it from somewhere.

"all-encompassing" (preferable short) are more useful than a single translation. Sometimes a single translation is enough though. When adding an entry, we should ask ourself, what should we write so it will be the most uself to ourself when
we meet that word again (and to others).
We will clarify this among other things in a FAQ soon.

You can only edit you own entries. If you think an entry from someone else can be improved, just create a new one, that's OK. Most favorited entries will be listed at the top.

Good to know that http://www.catalandictionary.org/ uses CC BY-SA, same license as the CEDICT.

Crush wrote:

EDIT: Also, what about a way to edit texts we've already added? A lot of my texts have odd line breaks in them, i suppose that's why i was having trouble with words spanning lines. Perhaps a way to share the texts in "My
Texts" would be nice, too. To me, it seems to make sense that EVERY text i add should go into my texts, perhaps you could have "My Texts" with a "Private Texts" subsection? I don't know if it was there before or not, but i do like being able to
select "my texts" in the reading section :)

At the moment, you can edit a text you have added by entering again the URL on the "add" page. (http://www.lingocracy.com/a/add-text) but we will make it simpler soon.
You are right, the currently UI about shared and private texts is not very clear. We will simplify it soon.
At the moment, only texts with an URL (added by URL) can be shared. Sharing texts from My Texts is not possible yet, but this may change at later point.

Crush wrote:

First off, i just noticed that you can now switch between your languages. Cool! However, on the page to change languages, the text seems to be cut off:

Quote:

We strongly recommend using English as dictionary (more entries and better quality). Of course, if

Also, why does Catalan use the Canadian flag? And when i want to change languages, there's an X that appears. I assume that's to remove the language from the list, which is a nice feature too. However, for me (Firefox) it shows up beneath
the language and is impossible to click since as soon as i go down to click on it it disappears when the mouse leaves the highlighted option.

Now that it's easier to change languages, i'd like to figure out how we can start implementing some more dictionaries for the languages i'm studying :)

The new process for adding texts is great, too!

Thanks!!
- Typo corrected
- Sorry for the language/flag error. We are working on that.
- "Remove language" button for Firefox should work now. (We previously only tested on Chrome)
- I'm glad you like the new language switcher :)

Edited by Yaan on 18 February 2014 at 7:45pm
1 person has voted this message useful

Henkkles
Triglot
Senior Member
Finland
Joined 4055 days ago
544 posts - 1141 votes

Speaks: Finnish*, English, Swedish
Studies: Russian

Message 32 of 134

18 February 2014 at 7:47pm | IP Logged

The new language switcher is great, it just displays the wrong flag for Estonian. You might want to change that.

I would also be forever grateful if you could add Northern Saami, the problem being that there are no dictionaries with English as the other language, but there are fantastic online dictionaries with Norwegian and Finnish bases that can be found here and here. The former actually enables extraction of morphological data, as in which is the base form of the word and which inflected form the word is in.

This is what their "about" page says:

Open-Source

Everything here is based on open-source tools. Do you have an idea for an app or research project that needs linguistic resources? Visit Giellatekno's website for information on how to get access to lexicon files, and morphological and syntactic analyzers.

I understand that this is likely a very low priority thanks to there not even being a Google translate but I figured that if I express my wish then at least it's noted.

Edited by Henkkles on 18 February 2014 at 7:53pm

2 persons have voted this message useful

This discussion contains 134 messages over 17 pages: << Prev 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 Next >>

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 2.2656 seconds.