1) JAPIO does not use MT for its abstracts.
2) Patent applicants in Japan supply an abstract with
the application and JAPIO apparently uses these as is.
(This is what they did with my patent applications.)
I highly recommend Derwent and CAS as they put
substantial intelligence into their abstracting. The deeper
I get into providing Japanese patent information, the more
I am in awe of the value and volume of work that Derwent
and CAS have done.
For a detailed, but somewhat dated description on databases
for Japanese patents, please see my paper "Using Online and
CDROM Patent Databases to Obtain Japanese Science
and Technology Information" at
http://www.intlscience.com/JapanPatentDatabases/intro.htm
Leland Ness wrote:
> This is a question for Alan, although I thought the response might be
> interesting to all the group, so I am posting it here. Now that my
> hopes of being able to machine-translate the Japanese patents that I
> need have been dashed, I have begun looking again at PAJ. In many of
> the abstracts I look at the abstract itself is unintelligible. The
> words are English, but the sentences (when such exist) often make no
> sense at all. I have spoken to several others who have encountered
> the same problem. What causes this? Is it: (a) the original
> (Japanese) abstracts are ambiguous(b) the MT program the JPO uses is
> so abysmal(c) the translation from Japanese to English is so complex
> that any MT program will yield the same result.(d) some problem I
> haven't thought of Thanks Lee Ness
-- Alan Engel ISTA, Inc. ConvertedKokai(tm) machine translations of Japanese patents http://www.intlscience.com