Relevance in Dialogue: Is Less More? An Empirical Comparison of Existing Metrics, and a Novel Simple Metric
This work addresses the challenge of reliable dialogue evaluation for researchers and practitioners, though it is incremental as it builds on existing metrics.
The authors tackled the problem of evaluating dialogue relevance metrics, finding that existing ones are dataset-dependent and poorly correlate with human scores; they proposed a novel simple metric that achieves state-of-the-art performance on the HUMOD dataset with a 37%-66% reduction in dataset sensitivity.
In this work, we evaluate various existing dialogue relevance metrics, find strong dependency on the dataset, often with poor correlation with human scores of relevance, and propose modifications to reduce data requirements and domain sensitivity while improving correlation. Our proposed metric achieves state-of-the-art performance on the HUMOD dataset while reducing measured sensitivity to dataset by 37%-66%. We achieve this without fine-tuning a pretrained language model, and using only 3,750 unannotated human dialogues and a single negative example. Despite these limitations, we demonstrate competitive performance on four datasets from different domains. Our code, including our metric and experiments, is open sourced.