Paper: OSU-2: Generating Referring Expressions with a Maximum Entropy Classifier

ACL ID W08-1130
Title OSU-2: Generating Referring Expressions with a Maximum Entropy Classifier
Venue International Conference on Natural Language Generation
Session Main Conference
Year 2008
Authors

Selection of natural-sounding referring ex- pressions is useful in text generation and in- formation summarization (Kan et al., 2001). We use discourse-level feature predicates in a maximum entropy classifier (Berger et al., 1996) with binary and n-class classification to select referring expressions from a list. We find that while mention-type n-class classifi- cation produces higher accuracy of type, bi- nary classification of individual referring ex- pressions helps to avoid use of awkward refer- ring expressions.