1Efficient Feedback-based Feature Learning for Blog Distillation as a Terabyte Challengeinformation, we first discover more (non-heuristic) feature candidates, including unigram features as well as some word...
2Extracting and Ranking Product Features in Opinion Documentsextracted feature candidates to improve the precision of the top-ranked candidates. We rank feature candidates by feature...
3Automatic Discovery of Feature Sets for Dependency Parsingenable a comparison of the scores. We evaluated the feature candidates on a development set using the labeled and unlabeled...