tech,0-1-N01-1003,ak </term> and <term> key prediction </term> . <term> Sentence planning </term> is a set of inter-related but distinct
tech,15-1-N01-1003,ak but distinct tasks , one of which is <term> sentence scoping </term> , i.e. the choice of <term> syntactic
other,22-1-N01-1003,ak scoping </term> , i.e. the choice of <term> syntactic structure </term> for <term> elementary speech acts </term>
other,25-1-N01-1003,ak <term> syntactic structure </term> for <term> elementary speech acts </term> and the decision of how to combine
other,40-1-N01-1003,ak how to combine them into one or more <term> sentences </term> . In this paper , we present <term>
tool,6-2-N01-1003,ak </term> . In this paper , we present <term> SPoT </term> , a <term> sentence planner </term> ,
tech,9-2-N01-1003,ak paper , we present <term> SPoT </term> , a <term> sentence planner </term> , and a new methodology for automatically
tool,19-2-N01-1003,ak methodology for automatically training <term> SPoT </term> on the basis of <term> feedback </term>
model,24-2-N01-1003,ak training <term> SPoT </term> on the basis of <term> feedback </term> provided by <term> human judges </term>
other,27-2-N01-1003,ak of <term> feedback </term> provided by <term> human judges </term> . We reconceptualize the task into
tech,6-4-N01-1003,ak distinct phases . First , a very simple , <term> randomized sentence-plan-generator ( SPG ) </term> generates a potentially large list
other,18-4-N01-1003,ak potentially large list of possible <term> sentence plans </term> for a given <term> text-plan input </term>
other,23-4-N01-1003,ak <term> sentence plans </term> for a given <term> text-plan input </term> . Second , the <term> sentence-plan-ranker
tech,3-5-N01-1003,ak text-plan input </term> . Second , the <term> sentence-plan-ranker ( SPR ) </term> ranks the list of <term> output sentence
other,11-5-N01-1003,ak sentence-plan-ranker ( SPR ) </term> ranks the list of <term> output sentence plans </term> , and then selects the <term> top-ranked
other,19-5-N01-1003,ak plans </term> , and then selects the <term> top-ranked plan </term> . The <term> SPR </term> uses <term> ranking
tech,1-6-N01-1003,ak the <term> top-ranked plan </term> . The <term> SPR </term> uses <term> ranking rules </term> automatically
model,3-6-N01-1003,ak plan </term> . The <term> SPR </term> uses <term> ranking rules </term> automatically learned from <term> training
lr,8-6-N01-1003,ak rules </term> automatically learned from <term> training data </term> . We show that the trained <term> SPR
tech,5-7-N01-1003,ak data </term> . We show that the trained <term> SPR </term> learns to select a <term> sentence
skrij podrobnosti