Introducing Chinese Text Analyser

[murrayjames]

Hi Imron,

 

Ask and you shall receive! Thanks for the improvements. There no need to uninstall the old version before updating, right?

 

By the way, I haven't forgot about my review. Waiting for 1.0, if that's ok -- I'll write in depth about the release version.

 

Does "Clear Known Words" have an "Are you sure"-type dialog box? I want to try it and see, but am afraid to lose my word history :-)

[imron]

No need to uninstall previous versions, and installing over the top will keep all your old settings.

Uninstalling on the other hand will delete everything. I don't like programs that leave things behind when uninstalled so CTA cleans up after itself (assuming no bugs).

Clear known words does have a dialog asking for confirmation.

[Angelina]

Waiting for the OS X version to come out 

[realmayo]

Is there any way to stop it automatically 'unknowing' a word whose definition I look up? I find it a bit patronising.

[imron]

No, and this is by design.  Its purpose is to force you to realise words that you aren't 100% confident with.

 

If you export these words and set 'mark exported words as known' then it will mark them as known.

[realmayo]

Well, if I wasn't 100% confident I wouldn't be checking. So, the point about not being 100% confident, is already realised.

 

If the software was called Chinese Text Reader With Teacher Watching Over You I might understand this horrible feature.

[realmayo]

Sorry if that sounds too grumpy! It's just, I can understand when software doesn't do what you want it to do because it's been badly designed. But in this cse, it won't do what I want it to do because someone else thinks they know best. Like, if someone said there's no way you can learn 20 new words a day so Anki or whatever wouldn't let you learn more than that.

[imron]

I'll forgive you your grumpiness if you forgive me my patronising (unfortunately this post is probably going to come across as patronising even though I don't mean it to be).   With that being said....

 

CTA is designed to reflect reality.

 

It's also designed to stop people coasting by on almost known vocabulary and help them move away from using a dictionary as a crutch.

 

Rather than asking you if you know a word, it looks at your actions, and looking up a word in the dictionary is an implicit acknowledgement that you didn't know that word well enough to read it (despite how painful admitting such a fact can be).

 

By marking such words as unknown, CTA provides an easy way to export those words for further study and drilling with the hope that you won't need a dictionary the next time. 

 

If you're not interested in learning those words further, leaving it marked as unknown is the more accurate depiction of the current state of events because if you haven't put in some mindful effort to learn the word you'll likely still be needing the dictionary for that word in one week's time.

 

I agree that this can be a little bit frustrating, especially if you feel you know the word and just wanted to look it up to check, but that's the whole point!  CTA is trying to spur you to learn those words properly so that next time you won't need to check them, you'll know.

 

I will consider the above name change before considering changing this feature

 

. Like, if someone said there's no way you can learn 20 new words a day so Anki or whatever wouldn't let you learn more than that.

But it's not, because CTA is not putting any sort of limitation on your learning.  It's just pointing out the current state of affairs.  You might not like that state of affairs, but CTA is just the messenger.  The real solution is to spend a bit more time learning vocabulary.

[realmayo]

I do understand the logic behind what you're saying. However I'm not using the programme as a reader, but as a tool to analyse Chinese text and see what vocab it contains that I haven't learned.

 

The next step is deciding whether an unlearned word might be worth learning. Sometimes I'll look at the sentence in which one of those unlearned words occurs: the context is helpful to deciding whether to learn it (and also to see if it's just separate characters that have been incorrectly segmented). And in that sentence there might be a learned word that I want to look up. It could be that I've forgotten it. But it could be that I'm just a teeny bit unsure, or want to double check one of the tones, that kind of thing.

 

I'm never going to remember *everything* I've learned. But nor am I going to immediately *relearn* everything that I've forgotten. I think that's impactical, to make it a blanket rule like that.

 

Anyway, I'll simply stop using the look-up function. As I say, it rankles a little bit, like 'teacher knows best'. Personally, I read on Pleco reader and that stores a list of any words looked-up, which I regularly scan for ones that are worth 'relearning'. So we agree about the importance of that. But not about whether I can be trusted or not!

 

And the programme is of course amazingly useful even without the look-up function so thanks again for producing it.

[imron]

But it could be that I'm just a teeny bit unsure, or want to double check one of the tones, that kind of thing.

These are the most dangerous words to look up with a popup dictionary because there's a tendency to look them up to check, see that it was what you thought, and then move on because you were 'correct', without spending any time thinking about why you needed to check in the first place and reinforcing some memories so you won't need to check the next time you encounter the word.

 

Anyway, I'll simply stop using the look-up function

Actually this was going to be one of my suggestions, and in fact you'll remember from earlier versions of the program that such a feature didn't even exist.  One of the goals of Chinese Text Analyser is to get people to move away from relying on popup dictionaries so in some ways you could say it is meeting its intended purpose.

 

Now you might say 'hah no it's not because I'm still going to go away and look it up on Pleco', but this is extra brain power and extra thinking about the word.  You've probably had to write the character out, or look it up by its pinyin, which will all act as minor reinforcements, and as you said Pleco keeps a list of looked up words you can always scan back later for learning, and all of this is the purpose of the 'mark looked up words as unknown' feature - to encourage you to get these words into your external study process.

[realmayo]

These are the most dangerous words to look up with a popup dictionary

 

So, I won't look them up at all, and will remain slightly unsure about them until I see them elsewhere. Which doesn't seem an improvement to me but anyway....

 

Yep basically I'll use the programme for what I initially thought it was designed for, import text, export unknown vocab. It's excellent at that. But not as a reader. For that I have Pleco Reader, which is ideal.

[Angelina]

I am using Pleco Reader on my Android devices. 

 

Imron, you might want to hurry up with that OS X version. I can't use the Pleco Reader on my Macbook. Hint hint.

 

I am using the Kindle app when reading books in Chinese on my Macbook. This is better than Pleco because you can manually select a two-character word and then click for the dictionary entry. A monolingual dictionary is available there and I now prefer to use it because it is much better to continue thinking in Chinese and not be disrupted by English. 

 

The Pleco Reader automatically decides what is a word and it is sometimes wrong. There was a thread about this before. 

 

 

Edited:

 

See post 196

[Angelina]

http://www.chinese-forums.com/index.php?/topic/41432-how-to-figure-out-if-a-character-is-used-as-a-word/

 

Here it is

[imron]

I am working on the OSX version as we speak, however it will have the same problems you mention with Pleco reader.  Chinese word segmentation is an imprecise and inexact science.  CTA also doesn't have a Chinese dictionary as yet.

[Angelina]

Does the Chinese Text Analyser at least have the option to manually decide what a word is? Pleco Reader doesn't. Kindle for Mac does. 

 

Edited:

See post below, #196

[Yadang]

 

The Pleco Reader automatically decides what is a word and it is sometimes wrong.

 

I thought you could manually change this. If you touch the screen to select a word it will automatically segment it, but if it segments out a word at the wrong place, you can use the arrows below to select individual characters (there are two sets of arrows, the ones on the right will segment and select words, but the ones on the left will allow you to select as many or as few characters as you want. If the characters match a dictionary entry, it will pop up). Have you tried that, or is that not what your talking about?

 

 

A monolingual dictionary is available there and I now prefer to use it because it is much better to continue thinking in Chinese and not be disrupted by English.

There are of course paid mono-lingual dictionaries you can buy on Pleco, or if you can read traditional characters, you can also use this free one from Taiwan.

[Yadang]

@Imron Any ETA's for Linux? And are you going to consider making an Android version?

[Angelina]

Thanks Yadang, I didn't know you could select characters using the arrows below. I thought they served only to move to the next character.

Thanks for the free dictionary too. 

[Yadang]

No problem! I hope you enjoy! However I am merely the messenger - alex_hk90 from Pleco Forums is the one who did all the work!

[imron]

No ETA on the Linux version, but it will be after the OSX one and mostly likely after (and definitely not before) the corpus feature.

I've thought about mobile, but anything in that space is still a long way off.

@angelina, you can mark what is a word in CTA, but currently this is very limited.