Re: Corpora: US company claims patent on common machine

Job vanZuijlen (zuijlen@ibm.net)
Mon, 10 Nov 1997 14:40:15 -0500

The fact that this patent has been granted at all, shows once more that
patent investigators don't do a very good job. At the time the patent
was filed (Febr. 1987), I was part of the research team at BSO in the
Netherlands that worked on a machine-translation project called DLT
(Distributed Language Translation) which used Esperanto as an
intermediate language, much in the same way as the patent claims it
does.
A comprehensive feasibility study for this project was completed in
1983. The study was publicly available at the time and also referenced
in other publications, so any patent investigator could have known about
it. The initial idea was proposed around 1979 within BSO, but, at the
time was not considered original enough to grant a patent. I am not a
patent lawyer, but if it can be clearly shown that something that is
patented did already exist, I don't see how anyone could be sued.
Unfortunately, I have seen quite a number of patents being granted for
ideas that are almost common knowledge within the CL or other
computational
communities.

Job van Zuijlen