A Norwegian-Portuguese DISPARA prototype
The present prototype was devised under the scope of the projects: Linguateca (more specifically under the COMPARA/DISPARA project) and Squirrel, Corpus based language technology for computer-assisted learning of Nordic languages.
This demo has two aims:
- Demonstrate the feasibility of using DISPARA for other language pairs
- Demonstrate how to use revised and raw translations to teach one language to natives of the other
Notes on the DISPARA adaptation
DISPARA refers to the whole system of putting parallel corpora on the Web.
Adapting DISPARA means basically the following:
- Using the previous programs to create the corpora DNPM_PORTM, DNPM_NOR and the corpora SAM_POR SAM_NOB and SAM_NOM, aligned as to UA
- Creating a basic "void" CGI interface to be customized automatically, against the more complex program that has more options: As parameters, it receives the corpus name and its description string, the monitoring file and the name(s) of the possible alignment attributes.
- Instantiate this void interface with the present corpora.
- Create a new HTML interface, the present one in this case.
Notes on the use of several corpora simultaneously for teaching
What we want to point out here is the possible use of aligned text and its correction in addition to the original.
So, one may look at "original" text (text written by a native), text rendered by a non-native and/or its subsequent correction/edition by a native.
One can search in either three corpora separately, or look at pairs failure/correction, or look at correction/failure or look at corresponding text in the two languages.
This is something we could expect specialists to be able to make good use of, as well as inspire the development of programs that suggest better writing in a foreign language...
Unfortunately the SQUIRREL project is no longer active.
Last update: 28 November 2003.