CliCR: A Dataset of Clinical Case Reports for Machine Reading Comprehension
This provides a new benchmark for evaluating machine comprehension in healthcare, though it is incremental as it adapts existing methods to a specific domain.
The authors tackled the problem of machine reading comprehension in the medical domain by creating CliCR, a dataset of clinical case reports with 100,000 gap-filling queries, and found a 20% F1 performance gap between the best human and machine readers.
We present a new dataset for machine comprehension in the medical domain. Our dataset uses clinical case reports with around 100,000 gap-filling queries about these cases. We apply several baselines and state-of-the-art neural readers to the dataset, and observe a considerable gap in performance (20% F1) between the best human and machine readers. We analyze the skills required for successful answering and show how reader performance varies depending on the applicable skills. We find that inferences using domain knowledge and object tracking are the most frequently required skills, and that recognizing omitted information and spatio-temporal reasoning are the most difficult for the machines.