nlp - Identifying and comparing syntactic structure of questio-sentence -

nlp - Identifying and comparing syntactic structure of questio-sentence -

i getting question user , trying understand syntactically.

my goal identify exact question sentence user entered question. like

obama president of usa, wife?

so able apply anaphora resolution , his pointing obama , can convert above sentence

obama president of usa, obama wife?

but how can syntactically identify exact question sentence i.e. who obama wife? above entire question

i trying pylinkgrammar give 54 linkage above sentence, like

linkparser> linkage 54, cost vector = (unused=0 dis= 8.05 len=24) +------------------------------xp------------------------------+ +---------------------->wv---------------------->+ | +-------------------xx-------------------+-->wv->+---sis---+ | +----wd---+--ss--+--oum--+---mp--+-js+ +wq+--q-+ +ds**c+ | | | | | | | | | | | | | left-wall obama[!] is.v president.t of usa.l , is.v wife.n ?

what want defining pattern different question type w5h1, conjunction based question etc.

but dont find how write rule these pattern, suggestion , reference much appreciable?

you can seek extract different possible sub-questions (hypotheses) original text , test textual entailment between text , hypotheses. check out http://hltfbk.github.io/excitement-open-platform/#recognizing_textual_entailment

nlp analytics nltk semantics