Domain-Relevant Embeddings for Medical Question Similarity
This addresses the challenge of handling high volumes of medical questions for healthcare systems, though it is incremental as it builds on existing pre-training methods for a specific domain.
The paper tackled the problem of identifying similar medical questions online to improve answer efficiency, achieving an accuracy of 82.6% with a semi-supervised pre-training approach, outperforming other methods that scored below 78.7%.
The rate at which medical questions are asked online far exceeds the capacity of qualified people to answer them, and many of these questions are not unique. Identifying same-question pairs could enable questions to be answered more effectively. While many research efforts have focused on the problem of general question similarity for non-medical applications, these approaches do not generalize well to the medical domain, where medical expertise is often required to determine semantic similarity. In this paper, we show how a semi-supervised approach of pre-training a neural network on medical question-answer pairs is a particularly useful intermediate task for the ultimate goal of determining medical question similarity. While other pre-training tasks yield an accuracy below 78.7% on this task, our model achieves an accuracy of 82.6% with the same number of training examples, and an accuracy of 80.0% with a much smaller training set.