# Attention in Natural Language Processing

## TL;DR

This paper provides a comprehensive overview and taxonomy of attention mechanisms in NLP, categorizing various models based on input representation, compatibility, distribution, and multiplicity, highlighting ongoing research and challenges.

## Contribution

It introduces a unified model and taxonomy for attention architectures in NLP, systematically categorizing existing models and discussing future research directions.

## Key findings

- Proposes a taxonomy based on four key dimensions.
- Provides a unified framework for understanding attention models.
- Discusses how prior information can be integrated into attention mechanisms.

## Abstract

Attention is an increasingly popular mechanism used in a wide range of neural architectures. The mechanism itself has been realized in a variety of formats. However, because of the fast-paced advances in this domain, a systematic overview of attention is still missing. In this article, we define a unified model for attention architectures in natural language processing, with a focus on those designed to work with vector representations of the textual data. We propose a taxonomy of attention models according to four dimensions: the representation of the input, the compatibility function, the distribution function, and the multiplicity of the input and/or output. We present the examples of how prior information can be exploited in attention models and discuss ongoing research efforts and open challenges in the area, providing the first extensive categorization of the vast body of literature in this exciting domain.

---
Source: https://tomesphere.com/paper/1902.02181