I am really thankful to those who tried to answer my previous questions.I have another question which goes as follows:
I have a sentence as following
"The dog chased the cat."
This is what I get after parsing:
(ROOT (S (NP (DT The) (NN dog)) (VP (VBD chased) (NP (DT the) (NN cat))) (. .))) Now I need to make triples and also draw ontology. How to do it? Are there any tools available for this purpose but I need something which does triple formation and ontology drawing automatically.
asked 03 Nov '12, 09:45
IMO, you'd better use the dependency parsing technique instead of the one you're doing now which is called the constituency parsing. This way you'd find it quite straightforward to get the triples from a sentence. Say, I have the following sentence:
The dependency graph for the sentence is the following:
And guess what? Yes, the results are already in triples. You just need to annotate them with terms from some NLP (Natural Language Processing) ontology.
answered 03 Nov '12, 14:49
@fadirra, can you advise me, please.
I have a document that has been through NIF (NLP2RDF), providing me with an ontology of the document and its sentences. This combines with a number of owl ontologies (olia, Stanford, Penn), to define grammar classes (nsubj, dobj, etc) and instances (sentences and words).
Using Stanford's CoreNLP, I can obtain a dependency graph for each sentence. How do I relate the two? How do I turn nsubj(makes-8, Bell-1) into an asserted fact?
I need to assert that the sentence has an instance of the class nsubj with a head (Bell) and a dependent (makes).
answered 15 Aug, 08:09