SideLines, March 2006
In the March 2006 issue of Scientific American (pp. 70-73), Gary Stix has reviewed the history of machine translation and recent attempts by varous online services to use statistical methods and rule-based systems to translate between many different languages. Kevin Hendzel, a spokesman for the American Translators Asssociation is quoted on the subject. He does not seem overly concerned with the competition...
As an instant test, and for the amusement of our readers, we have submitted the phrase
"Machine Translation"
to a rule-based Google/Systran translation from English to Portugese and from Portugese back to English.
Here is the result:
"Translation De Máquina"
Perhaps the real value of machine translation in the future will be the creation a universal hybrid language (??)
Contact The Research Cooperative (NZ): info (at) researchco-op (dot) co.nz )
Tuesday, March 07, 2006
Absurdly popular catch-phrases?
Recently, at our main website (The Research Cooperative NZ), three recent postings have been attracting unusually large number of visitors.
Is this an absurdity, some kind of scam, or simply the result of using good catch-phrases?
Our Co-op member has used phrases that clearly identify the services offered, as follows:
1. English-Chinese translation (464 visits in one month)
2. English editing (302 visits in one month)
3. English proofreading (192 visits in one month)
As a rough test of why these offers have attracted attention, I present results (hit numbers) for the following exact search phrases, using Google:
First set
English-Chinese translation - 52,600
English-Chinese - 2,160,000
English to Chinese translation - 138,000
English to Chinese - 1,900,000
Chinese-English translation -- 66,300
Chinese-English - 2,700,000
English-Mandarin - 228,000
English-Cantonese - 158,000
English-German translation - 1,250,000
English-German - 10,500,000
English-Spanish translation - 194,000
English-Spanish - 12,700,000
Maybe more people (especially English readers) are seeking translation from Chinese into English than from English into Chinese - and maybe it is better for translators and writers to avoid being very specific about which Chinese language they are working with, when designing a catch phrase, unless they are working with Cantonese (which I guess has a smaller speech community than Mandarin).
Also, there is more English-language communication about translation between European languages, than between English and Chinese, which is not surprising. It would be of interest to have comparable figures for Chinese-language areas of the Internet.
Second set
English editing - 55,200
Chinese editing - 809
i.e. Not many English readers seek to write in Chinese.
Third set
English proofreading - 25,100
Chinese proofreading - 305
i.e. Not many English readers seek editors of Chinese.
Comments? Please tell us - what do you think these numbers mean?
Is this an absurdity, some kind of scam, or simply the result of using good catch-phrases?
Our Co-op member has used phrases that clearly identify the services offered, as follows:
1. English-Chinese translation (464 visits in one month)
2. English editing (302 visits in one month)
3. English proofreading (192 visits in one month)
As a rough test of why these offers have attracted attention, I present results (hit numbers) for the following exact search phrases, using Google:
First set
English-Chinese translation - 52,600
English-Chinese - 2,160,000
English to Chinese translation - 138,000
English to Chinese - 1,900,000
Chinese-English translation -- 66,300
Chinese-English - 2,700,000
English-Mandarin - 228,000
English-Cantonese - 158,000
English-German translation - 1,250,000
English-German - 10,500,000
English-Spanish translation - 194,000
English-Spanish - 12,700,000
Maybe more people (especially English readers) are seeking translation from Chinese into English than from English into Chinese - and maybe it is better for translators and writers to avoid being very specific about which Chinese language they are working with, when designing a catch phrase, unless they are working with Cantonese (which I guess has a smaller speech community than Mandarin).
Also, there is more English-language communication about translation between European languages, than between English and Chinese, which is not surprising. It would be of interest to have comparable figures for Chinese-language areas of the Internet.
Second set
English editing - 55,200
Chinese editing - 809
i.e. Not many English readers seek to write in Chinese.
Third set
English proofreading - 25,100
Chinese proofreading - 305
i.e. Not many English readers seek editors of Chinese.
Comments? Please tell us - what do you think these numbers mean?
Subscribe to:
Posts (Atom)