Explaining Recurrent Neural Network Predictions in Sentiment Analysis
This work provides incremental improvements in interpretability for sentiment analysis, benefiting researchers and practitioners using RNNs.
The authors tackled the problem of explaining predictions in recurrent neural networks by extending Layer-wise Relevance Propagation (LRP) to handle multiplicative connections in LSTMs and GRUs, achieving better results than a gradient-based method on a five-class sentiment prediction task.
Recently, a technique called Layer-wise Relevance Propagation (LRP) was shown to deliver insightful explanations in the form of input space relevances for understanding feed-forward neural network classification decisions. In the present work, we extend the usage of LRP to recurrent neural networks. We propose a specific propagation rule applicable to multiplicative connections as they arise in recurrent network architectures such as LSTMs and GRUs. We apply our technique to a word-based bi-directional LSTM model on a five-class sentiment prediction task, and evaluate the resulting LRP relevances both qualitatively and quantitatively, obtaining better results than a gradient-based related method which was used in previous work.