Reputation: 41
Using Stanford Core NLP 3.7.0 to create Dependency tree in C#. Same sentence giving different output in my application and the online Parser demo
Sentence: Display Prime HomePage on page load.
Online demo output:
Universal dependencies
root(ROOT-0, Display-1)
compound(HomePage-3, Prime-2)
dobj(Display-1, HomePage-3)
case(load-6, on-4)
compound(load-6, page-5)
nmod(HomePage-3, load-6)
My C# application output:
Universal dependencies
compound(homepage-3, display-1)
amod(homepage-3, prime-2)
root(ROOT-0, homepage-3)
case(load-6, on-4)
compound(load-6, page-5)
nmod:on(homepage-3, load-6)
I am not getting the dobj which is available in Online demo output. Please advice.
Thank you.
Upvotes: 1
Views: 212
Reputation: 41
Following are the points i have noted, which may be helpful for others,
Stanford CoreNLP and Stanford Parser are not providing similar output. It differs for the same input sentence.
Stanford CoreNLP - Case sensitive. Getting different output if the sentence case is modified.
Upvotes: 1