This is probably caused by the incorrect tokenization. The POS Tagger also looks at parts of the word and that probably causes the different tags. Try to run this with "book." tokenized as "book .", and similar for website.
Jörn On Wed, Sep 6, 2017 at 1:35 PM, Gagneet Kaur <[email protected]> wrote: > Hi, > > I have a query regarding the Parse Tree structure of following two > sentences: > > *Sentence#1:* > > *"LanguageProcessing is a very huge and useful website."* > > *The Parse Tree is:* > > (TOP (S (NP (NN LanguageProcessing)) (VP (VBZ is) (NP (DT a) (ADJP (RB > very) (JJ huge) (CC and) (JJ useful)))) *(. website.)))* > > > *Sentence#2:* > > *"LanguageProcessing is a very huge and useful book."* > > *The Parse Tree is:* > > (TOP (S (NP (NN LanguageProcessing)) (VP (VBZ is) (NP (DT a) (ADJP (RB > very) (JJ huge) (CC and) (JJ useful)) *(NN book.)))))* > > > Considering that both the words '*website*' and '*book*' are nouns, there > seems to be a difference in the Parse Trees of both sentences. > *Parts of Speech* for word '*website*' in the first sentence above is > marked by a dot(.) while for the word '*book*' in the second sentence > above, it is marked as *NN* Part of Speech. > > Also, the *structure *of the trees gets *altered *due to this behavior. > > Could you please provide some explanation for the difference in the parsing > and signify the meaning of *dot(.)* before the word '*website'* in the > Parse Tree of the first Sentence. > > > Thanks, > Gagneet Kaur.
