other,1-2-H92-1060,bq linguistic analysis </term> fails . This <term> robust parsing capability </term> was achieved through minor extensions
other,44-3-H92-1060,bq semantic frame </term> encoding the full <term> meaning </term> of the <term> sentence </term> . We have
other,47-3-H92-1060,bq the full <term> meaning </term> of the <term> sentence </term> . We have assessed the degree of
tech,18-2-H92-1060,bq </term> already in place for the full <term> linguistic analysis component </term> . Robust <term> parsing </term> is applied
tech,13-5-H92-1060,bq </term> allowed us to answer many more <term> questions </term> correctly , as over a third of the
tech,22-5-H92-1060,bq correctly , as over a third of the <term> sentences </term> were not covered by the <term> grammar
tech,23-1-H92-1060,bq </term> , which allows it to answer a <term> question </term> when a full <term> linguistic analysis
lr-prod,12-6-H92-1060,bq performance of the <term> system </term> on the <term> February '92 test sentences </term> , and discuss some issues with regard
other,24-4-H92-1060,bq of robustly parsed vs. fully parsed <term> sentences </term> on the <term> October '91 dry-run test
other,39-3-H92-1060,bq gluing them together to obtain a single <term> semantic frame </term> encoding the full <term> meaning </term>
measure(ment),17-4-H92-1060,bq mechanism </term> through a breakdown of the <term> performance </term> of robustly parsed vs. fully parsed
tool,7-1-H92-1060,bq paper describes an extension to the <term> MIT ATIS ( Air Travel Information Service ) system </term> , which allows it to answer a <term>
tech,27-1-H92-1060,bq a <term> question </term> when a full <term> linguistic analysis </term> fails . This <term> robust parsing
lr-prod,27-4-H92-1060,bq parsed <term> sentences </term> on the <term> October '91 dry-run test set </term> . It was clear that the robust <term>
lr,28-5-H92-1060,bq sentences </term> were not covered by the <term> grammar </term> . We also report here on the performance
tool,9-6-H92-1060,bq report here on the performance of the <term> system </term> on the <term> February '92 test sentences
tech,6-5-H92-1060,bq </term> . It was clear that the robust <term> parser </term> allowed us to answer many more <term>
tech,25-6-H92-1060,bq discuss some issues with regard to the <term> evaluation methodology </term> . The <term> CSR ( Connected Speech
tech,10-4-H92-1060,bq the degree of success of the robust <term> parsing mechanism </term> through a breakdown of the <term> performance
other,27-3-H92-1060,bq parsing a set of <term> phrases </term> and <term> clauses </term> , and 2 ) gluing them together to
hide detail