DiSAN: Directional Self-Attention Network for RNN/CNN-Free Language Understanding

TL;DR

DiSAN introduces a novel directional self-attention mechanism that enables effective sentence encoding without RNNs or CNNs, achieving state-of-the-art results across multiple NLP benchmarks with improved efficiency.

Contribution

The paper presents a lightweight, RNN/CNN-free attention model called DiSAN that outperforms existing sentence encoding methods on various NLP tasks.

Findings

Outperforms RNN models in prediction quality and efficiency

Achieves best accuracy on SNLI dataset, improving by 1.02%

State-of-the-art results on multiple NLP benchmarks

Abstract

Recurrent neural nets (RNN) and convolutional neural nets (CNN) are widely used on NLP tasks to capture the long-term and local dependencies, respectively. Attention mechanisms have recently attracted enormous interest due to their highly parallelizable computation, significantly less training time, and flexibility in modeling dependencies. We propose a novel attention mechanism in which the attention between elements from input sequence(s) is directional and multi-dimensional (i.e., feature-wise). A light-weight neural net, "Directional Self-Attention Network (DiSAN)", is then proposed to learn sentence embedding, based solely on the proposed attention without any RNN/CNN structure. DiSAN is only composed of a directional self-attention with temporal order encoded, followed by a multi-dimensional attention that compresses the sequence into a vector representation. Despite its simple form, DiSAN outperforms complicated RNN models on both prediction quality and time efficiency. It achieves the best test accuracy among all sentence encoding methods and improves the most recent best result by 1.02% on the Stanford Natural Language Inference (SNLI) dataset, and shows state-of-the-art test accuracy on the Stanford Sentiment Treebank (SST), Multi-Genre natural language inference (MultiNLI), Sentences Involving Compositional Knowledge (SICK), Customer Review, MPQA, TREC question-type classification and Subjectivity (SUBJ) datasets.