USR: An Unsupervised and Reference Free Evaluation Metric for Dialog Generation.

Shikib Mehri Maxine Eskénazi

Published in: CoRR (2020)

Keyphrases

evaluation metrics
precision and recall
average precision
evaluation measures
semi supervised
unsupervised learning
evaluation framework
natural language
supervised learning
pairwise
learning to rank
information retrieval
decision trees
user interface
information extraction
co occurrence