What does the tag SBAR mean in Stanford’s parse-tree representation?

parsing nlp stanford-nlp parse-tree
user3054733 picture user3054733 · Dec 1, 2013 · Viewed 7.2k times · Source

When the Online Stanford Parser tool is fed this original sentence:

After she ate the cake, Emma visited Tony in his room.

It produces the following parse-tree representation as its output:

(ROOT
  (S
    (SBAR (IN After)
      (S
        (NP (PRP she))
        (VP (VBD ate)
          (NP (DT the) (NN cake)))))
    (, ,)
    (NP (NNP Emma))
    (VP (VBD visited)
      (NP
        (NP (NNP Tony))
        (PP (IN in)
          (NP (PRP$ his) (NN room)))))
    (. .)))

My questions are:

  1. What does the SBAR tag mean?
  2. Why are there two different S tags?
  3. What is the correct NLP parse-tree representation of this sentence?

Answer

user1906763 picture user1906763 · Jan 7, 2014

SBAR stands for Subordinate Clause (see here). In your case the subordinate clause starts with the subordinate conjunction After

Related questions

Stanford Parser and NLTK

Is it possible to use Stanford Parser in NLTK? (I am not talking about Stanford POS.)

Read More
How can I split a text into sentences using the Stanford parser?

How can I split a text or paragraph into sentences using Stanford parser? Is there any method that can extract sentences, such as getSentencesFromString() as it's provided for Ruby?

Read More
How to use Stanford parser

I downloaded the Stanford parser 2.0.5 and use Demo2.java source code that is in the package, but After I compile and run the program it has many errors. A part of my program is: public class testStanfordParser { /** Usage: ParserDemo2 [[grammar] …

Read More