Neural String Edit Distance
This provides a flexible framework for string processing tasks that balances performance and interpretability, though it is incremental in combining existing concepts.
The authors tackled the problem of string-pair matching and transduction by developing a neural string edit distance model that integrates learnable edit distance into neural networks. They achieved state-of-the-art performance on tasks like cognate detection and transliteration while offering a trade-off between accuracy and interpretability.
We propose the neural string edit distance model for string-pair matching and string transduction based on learnable string edit distance. We modify the original expectation-maximization learned edit distance algorithm into a differentiable loss function, allowing us to integrate it into a neural network providing a contextual representation of the input. We evaluate on cognate detection, transliteration, and grapheme-to-phoneme conversion, and show that we can trade off between performance and interpretability in a single framework. Using contextual representations, which are difficult to interpret, we match the performance of state-of-the-art string-pair matching models. Using static embeddings and a slightly different loss function, we force interpretability, at the expense of an accuracy drop.