Assessing agreement on classification tasks: the kappa statistic
Computational Linguistics
Dialogue actions for natural language interfaces
IJCAI'95 Proceedings of the 14th international joint conference on Artificial intelligence - Volume 2
| Hi-index | 0.00 |
This paper investigates tag-sets of coding schemes used for two-agent, task-oriented, problem-solving dialogue act annotation. The observed schemes are analyzed using criteria like sufficient documentation, easy transferability to other domains, and reliability of annotation. Tag-sets of schemes are compared, and common phenomena are stated. The aim of the paper is to provide an overview of currently available dialogue act coding schemes and to give scheme developers support in their scheme design.