USR: An Unsupervised and Reference Free Evaluation Metric for Dialog Generation.
Shikib MehriMaxine EskénaziPublished in: ACL (2020)
Keyphrases
- evaluation metrics
- precision and recall
- average precision
- learning to rank
- evaluation framework
- supervised learning
- unsupervised learning
- mixed initiative
- natural language
- active learning
- data sets
- evaluation measures
- benchmark datasets
- unsupervised manner
- collaborative filtering
- relational databases
- support vector
- information retrieval