Attention-based Conditioning Methods for External Knowledge Integration
This work addresses the challenge of enhancing RNNs with external knowledge for tasks like NLP, though it appears incremental as it builds on existing attention mechanisms.
The paper tackled the problem of integrating external knowledge into Recurrent Neural Networks by proposing lexicon features in the self-attention mechanism, resulting in consistent performance improvements across six benchmark datasets with minimal computational overhead.
In this paper, we present a novel approach for incorporating external knowledge in Recurrent Neural Networks (RNNs). We propose the integration of lexicon features into the self-attention mechanism of RNN-based architectures. This form of conditioning on the attention distribution, enforces the contribution of the most salient words for the task at hand. We introduce three methods, namely attentional concatenation, feature-based gating and affine transformation. Experiments on six benchmark datasets show the effectiveness of our methods. Attentional feature-based gating yields consistent performance improvement across tasks. Our approach is implemented as a simple add-on module for RNN-based models with minimal computational overhead and can be adapted to any deep neural architecture.