tech,1-1-N06-2009,ak its performance . State-of-the-art <term> Question Answering ( QA ) systems </term> are very sensitive to variations
other,17-1-N06-2009,ak to variations in the phrasing of an <term> information need </term> . Finding the preferred language
other,7-2-N06-2009,ak the preferred language for such a <term> need </term> is a valuable task . We investigate
tech,8-3-N06-2009,ak investigate that claim by adopting a simple <term> MT-based paraphrasing technique </term> and evaluating <term> QA system performance
other,13-3-N06-2009,ak paraphrasing technique </term> and evaluating <term> QA system performance </term> on paraphrased questions . We found
measure(ment),9-4-N06-2009,ak found a potential increase of 35 % in <term> MRR </term> with respect to the original question
skrij podrobnosti