AT deleted Our_NLP_system_is.tex  almost 9 years ago

Commit id: a96c8ebc9c8fe858a915751ec287fbd6c3bc853a

deletions | additions      

         

Our NLP system is a \textit{classifier}, i.e.\ a program that takes  a sentence (and a large knowledge base) and classifies it as either  true or false. The typical approach in such a scenario is building  a \textit{gold standard} dataset --- a set of questions, each annotated  with the correct answer (yes, no, unknown) and an accompanying knowledge  base that should, on its own, lend support to deciding these questions.%  \footnote{With a large enough knowledge base, this issue tends to vanish,  of course.}         

The_rest_of_the.tex  sectionQA_System_in_.tex  When_building_an_NLP.tex  Our_NLP_system_is.tex  These_questions_are_.tex  To_get_the_dataset.tex  The_main_criterium_f.tex