Multi-Zone Unit for Recurrent Neural Networks
This addresses a known bottleneck in RNNs for sequence learning problems, offering an incremental improvement over existing methods.
The paper tackles the difficulty of capturing complicated long-range dependencies in recurrent neural networks (RNNs) by introducing a Multi-zone Unit (MZU) with a transition function for multiple space composition, and demonstrates its superiority on character-level language modeling and aspect-based sentiment analysis tasks.
Recurrent neural networks (RNNs) have been widely used to deal with sequence learning problems. The input-dependent transition function, which folds new observations into hidden states to sequentially construct fixed-length representations of arbitrary-length sequences, plays a critical role in RNNs. Based on single space composition, transition functions in existing RNNs often have difficulty in capturing complicated long-range dependencies. In this paper, we introduce a new Multi-zone Unit (MZU) for RNNs. The key idea is to design a transition function that is capable of modeling multiple space composition. The MZU consists of three components: zone generation, zone composition, and zone aggregation. Experimental results on multiple datasets of the character-level language modeling task and the aspect-based sentiment analysis task demonstrate the superiority of the MZU.