Programs at the Text Laboratory
Diana Santos
List of programs
Here is a list of the programs available at the Text Laboratory, with links to user's guides (mostly in Norwegian):
Overview of the programs
Here we give an overview of the programs, according to
There are basically three kinds of programs:
- those running on
personal computers (IBM PCs or Macintoshs) which are available for
downloading from the Textlab disk (or in special cases of a network licence, which are run
from it)
- those running on Unix machines which must be invoked by logging on
to the users Unix account (for example, by using
telnet on PC's or MacX on Mac's)
- those following a client-server architecture, allowing users to remain in their small platforms, with the
style of user interfaces of their choice, while having a big
machine serving the results.
The first category comprises programs which are obviously smaller,
simpler and less powerful, but which are definitely worth while for
beginners. They are typically concordancing programs, or dictionary lookup
programs, most of which are documented at
http://www.hf.uio.no/tekstlab/bruksanvisning/index.html.
The second category requires much more computer proficiency on the
part of the users, but can be worth while if your requirements are
high,
or if you want to access particular resources, such as those created by international
projects
- WordNet
- The CHILDES database
- The CELEX database
NB! We have also licenses for some programs used in the projects currently
running at the Text laboratory, such as
- The Constraint Grammar parser of Lingsoft
- The Xerox statistical tagger
Although these are not for public access, it is conceivable that other
linguists use it if specifically hired by the Text Laboratory; or that
we can give demonstrations that may lead other projects to use them.
Finally the third kind of program/environment is fashionable nowadays,
and we expect to do some work in this paradigm at the Textlab pretty soon
(possibly in the autumn this year):
- the SARA client-server for the BNC
- simple CGI-based services to be accessed through the user's Web
browsers (e.g. Netscape)
In fact, working Web services (some more or less experimental) for the
systems we have here can already be found at several places:
Most programs we have either give you access to large bulks of
information:
- on-line dictionaries such as the OED,
- corpora such as the BNC, or PHI beta-coded texts for ancient texts
- databases, such as the CELEX database, or WordNet
Or are meant to apply to your own things and perform tasks on them:
- concordancers and browsers, such as MonoConc, Conc
- automatic translators, such as Nyno (See a detailed critique, in Norwegian, of
this program in a former Tekstlab bulletin)
- parallel browsers, such as ParaConc
A not so easy category are those programs that sometimes require so
much processing that actually they can be applied to more than one
corpus but in practice require these corpus to have been translated
into their internal format by specialists: this is the case with Tact,
Wordcruncher and Xkwic. You can do it yourself - just read the manual;
but, in practice, most people only use the texts that are already in that form.
Or still are meant for you to create resources easily.
- database programs, such as FileMaker Pro
- grammar development environments, such as the Constraint Grammar
Development environment
- tagger builders, such as the Xerox toolkit
Please note that this page is still under development.
Last modified 23 February 1999 by AN. Contact us.
Enhet: Det humanistiske fakultet, webmaster@hf.uio.no
Dokument opprettet: 17.07.2001, endret: 17.07.2001, verifisert: 17.07.2001