Reputation defence is a form of persuasive tactic that is used in various social settings especially in political situations. Detection of reputation defence strategy is a novel task that could help in argument reasoning. Here, we propose an approach to automatically label training data for reputation defence strategies. We experimented with about 14,000 pairs of questions and answers from the Canadian Parliament, and automatically created a corpus of questions and answers annotated with reputation defence strategies. We further assess the quality of the automatically labeled data.
@InProceedings{NADERI18.18, author = {Nona Naderi and Graeme Hirst}, title = {Automatically labeled data generation for classification of reputation defence strategies}, booktitle = {Proceedings of the Eleventh International Conference on Language Resources and Evaluation (LREC 2018)}, year = {2018}, month = {may}, date = {7-12}, location = {Miyazaki, Japan}, editor = {Darja Fišer and Maria Eskevich and Franciska de Jong}, publisher = {European Language Resources Association (ELRA)}, address = {Paris, France}, isbn = {979-10-95546-02-3}, language = {english} }
Summary of the paper