Evaluating neural network explanation methods using hybrid documents and morphological agreement
This provides a comprehensive evaluation framework for explanation methods in NLP, addressing the need for interpretability in DNNs, though it is incremental as it builds on existing methods like LIME.
The paper tackled the problem of evaluating post hoc explanation methods for deep neural networks in NLP by designing two novel evaluation paradigms for small and large context problems, and introduced LIMSSE, an NLP-inspired method, finding that LIMSSE, LRP, and DeepLIFT are the most effective.
The behavior of deep neural networks (DNNs) is hard to understand. This makes it necessary to explore post hoc explanation methods. We conduct the first comprehensive evaluation of explanation methods for NLP. To this end, we design two novel evaluation paradigms that cover two important classes of NLP problems: small context and large context problems. Both paradigms require no manual annotation and are therefore broadly applicable. We also introduce LIMSSE, an explanation method inspired by LIME that is designed for NLP. We show empirically that LIMSSE, LRP and DeepLIFT are the most effective explanation methods and recommend them for explaining DNNs in NLP.