I need to parse or tokenize English sentences. Is there any NLP toolkit in Prolog? Thanks.
Asked
Active
Viewed 1,357 times
1 Answers
5
Attempto is IMO your best option. The parser is implemented in SWI-Prolog, and has other tools mainly implemented in Java.
Lower level and a little outdated, from SWI-Prolog links page, there is ProNTO. I'm sorry, I've never tried any of these components.
Prolog was born as a natural language processor: but (maybe cause it evolved as a general purpose language) today is not the preferred choice for the task. The Wikipedia page, to be true very incomplete, doesn't report any Prolog toolkit.

CapelliC
- 59,646
- 5
- 47
- 90
-
Prolog is not the preferred choice because empirical and statistical methods have been dominant in NLP for almost two decades. Prolog doesn't support those out of the box and in fact its arithmetic is a drag. – Fred Foo Dec 02 '11 at 13:59
-
you are right, altough ProNTO has statisticals tools, this approach seems less appealing in Prolog – CapelliC Dec 06 '11 at 19:36
-
"Prolog was born as a natural language processor: but today is not the preferred choice for the task" Which language is? – Nikhil Prabhu Jan 05 '17 at 12:31
-
@NikhilPrabhu: maybe [GF](http://www.grammaticalframework.org/) but it also isn't listed on the wikipedia page – CapelliC Jan 05 '17 at 12:38