At this week’s lab meeting, Koustuv Sinha will present on Learning an Unreferenced Metric for Online Dialogue Evaluation (ACL, 2020).
- Wednesday, April 29th, at 14:30 UTC-4 (note the time change again!)
- Via Zoom. Contact Emily for details.
Evaluating the quality of a dialogue interaction between two agents is a difficult task, especially in open-domain chit-chat style dialogue. There have been recent efforts to develop automatic dialogue evaluation metrics, but most of them do not generalize to unseen datasets and/or need a human-generated reference response during inference, making it infeasible for online evaluation. Here, we propose an unreferenced automated evaluation metric that uses large pre-trained language models to extract latent representations of utterances, and leverages the temporal transitions that exist between them. We show that our model achieves higher correlation with human annotations in an online setting, while not requiring true responses for comparison during inference.
Koustuv is a second year PhD candidate at McGill University / Mila. His research interests lies in the intersection of open-ended dialog and systematic generalization in natural language and graph structured data.