[asialex] RE: [euralex] Josà Aguirre's metalexicography

  • From: "Gilles-Maurice de Schryver" <gillesmaurice.deschryver@xxxxxxxx>
  • To: "'David Joffe'" <david.joffe@xxxxxxxxxxxxxx>, 'Josà Aguirre' <jaguirreuk@xxxxxxxxx>
  • Date: Thu, 8 Nov 2012 23:20:01 +0100

Thanks David -- perhaps we ought to declare interest here: we're (also) in the 
language technology business (but not for CJK, yet) ... 

 

So, based on the video at the link below, I'd say the contest has been won by 
the dictionary of the future already (unless Josà Aguirre's handwriting is that 
of a medical doctor: sorry wanted a lighter note ;-). 

 

Perhaps I should also point out that more than just single Kanji characters are 
recognized at a time here (which was the initial challenge): the dictionary of 
the future recognizes full meaningful chunks, to take one from the videoéå 
âflashâ.

 

That's thus 10 for the electronic dictionary, zero for the paper dictionary.

 

Let's bring on the next challenge, please!

 

This is not a joke, this is not about techies having fun, colleagues, what we 
mean when we say that the "second revolution" in our field has arrived, is 
exactly examples like this. Leave the paper world behind, and start viewing 
lexicography in the digital age. Coming up with new solutions to the age-old 
look-up problems in Chinese and Japanese dictionaries is one of them.

 

All best,

Gilles-Maurice.

 

 

-----Original Message-----
From: David Joffe [mailto:david.joffe@xxxxxxxxxxxxxx] 
Sent: donderdag 8 november 2012 22:39
To: 'Josà Aguirre'; gillesmaurice.deschryver@xxxxxxxx
Cc: euralex@xxxxxxxxxxxxx; asialex@xxxxxxxxxxxxx; afrilex@xxxxxxxxxxxxx; 
DSNA@xxxxxxxxxxxxxxx; lexicographylist@xxxxxxxxxxxxxxx; ishll@xxxxxxxxxxxxxx; 
lexicografie@xxxxxxxxxxxxxxxx
Subject: Re: [euralex] Josà Aguirre's metalexicography

 

On 8 Nov 2012 at 21:52, Gilles-Maurice de Schryver wrote:

 

>     > if I jot down a random Chinese character on a piece of paper for both 
> of us to look up, 10 times 

>     out of 10 I will find it in a paper dictionary before you do in your 
> digital dictionary.

 

>     So here's a nice challenge for the CJK gurus! If Jack Halpern's 

> tools can't already beat you on this, let this be the Deep Blue 

> lexicographic equivalent. Jack!?

 

If I am not mistaken, digital solutions for this problem have already begun to 
be implemented, e.g.:

 

 <http://www.techinasia.com/pleco-dictionary-android/> 
http://www.techinasia.com/pleco-dictionary-android/

 

Basically, point your smartphone camera at a character, it runs it through OCR, 
and performs a dictionary search for you. I'm sure it's not perfect, but it's 
first-generation technology ... I don't know how this particular implementation 
would perform in a '10 attempts' 

'paper vs electronic' contest, but I expect these methods would improve a lot 
in the next 10 years:

 

 

".. the Android iteration of Pleco dictionary has today gone gold, and now 
finds a home in the Android Market. It comes with OCR abilities so that it can 
scan and âreadâ Chinese characters using your smartphoneâs camera, handwriting 
support, voice recognition, and numerous dictionary options.

 

Its range of features means that it can be used by the most casual of tourists 
who might want to scan a menu whilst visiting China, to the most studious of 
students of the Chinese language who might need to add specialist dictionaries 
and make flashcards"

 

 

 

- David

 

Other related posts: