Syntactic Structures of Sentences from Large Corpora 
Emmanuel Giguet and Jacques Vergne 
GREYC -- CNRS UPRESA 6072 -- Universit~ de Caen 
14032 Caen cedex -- France 
{Emmanuel.Giguet, Jacques.Vergne}@info.unicaen.fr 
1 The Demonstration 
The demonstration will consist in displaying the 
syntactic structure of sentences from novels, scien- 
tific texts, newspapers, ... The syntactic structures 
are computed by our syntactic parser ad the out- 
put shows in a human-friendly graphic interface (1) 
word features (as computed by POS tagger) (2) non- 
recursive phrases (as computed by shallow parser) 
and (3) their relations (the functional structure). It 
will be a closed demo. The outputs will be split into 
sets in order to focus on the resolution of different 
kinds of problems (i.e. coordination, long distance 
subject-verb relation, preposition/determiner reso- 
lution). A large scale corpus will be available to 
prove that the results are not hand-made and open 
demos will be possible during informal demo ses- 
sions. 
2 Demo Sessions 
Schedule of the intended demo sessions: 
Tuesday April 1 12.30-13.30 
16.30-17.30 
Wednesday April 2 10.00-11.00 
15.30-17.00 
Thursday April 3 10.30-12.00 
15.00-16.00 
16.30-17.00 
Other informal sessions can be scheduled in 
order to allow people to parse their own sentences 
or texts. 
3 The Graphic Output 
Our viewer, developped with Java, allows to display 
in a graphical way (see Fig. 1) the dependency tree 
between non-recursive phrases and the tags of words. Figure 1: "A GRAPHIC OUTPUT 
Figure 2: THE CONTROL PANEL 
4 The Control Panel 
The control panel (see Fig. 2) still in progress al- 
lows the choice of the corpus, the sentence to be 
displayed. Filters can be applied in order to show or 
to hide specified dependency and coordination re- 
lations. The tags of the words can also be shown 
or hidden. This principle allows to concentrate 
on word-level or on dependency-level and allows to 
check precisely the behavior of our parser on spe- 
cific relation types. A trace mode enables the step 
by step display of the construction of the analysis. 
In a near future, a graphic comparison between 
two outputs will be available to control the modifi- 
cation of the rule base and to point out easily the 
differences. This comparison between two outputs 
can also be used to compare an output and an ex- 
pected output. 
5 Web Site Opening 
After this conference, the same demonstration will 
be available on Internet at the following URL. 
http://www.info.unicaen.fr/~iguet 
This website will required a browser running 
Java and is platform independent (It has been 
tested with Netscape on Linux, Solaris2.5, MacOS, 
Windows NT and Windows 95). 
2 
6 Getting Annotated Corpora 
People who are interested in getting annotated cor- 
pora are invited to contact the authors. 
7 References 
No references are available yet. 
