Adversarial Deep Averaging Networks for Cross-Lingual Sentiment Classification

Abstract

In recent years great success has been achieved in sentiment classification for English, thanks in part to the availability of copious annotated resources. Unfortunately, most languages do not enjoy such an abundance of labeled data. To tackle the sentiment classification problem in low-resource languages without adequate annotated data, we propose an Adversarial Deep Averaging Network (ADAN) to transfer the knowledge learned from labeled data on a resource-rich source language to low-resource languages where only unlabeled data exists. ADAN has two discriminative branches: a sentiment classifier and an adversarial language discriminator. Both branches take input from a shared feature extractor to learn hidden representations that are simultaneously indicative for the classification task and invariant across languages. Experiments on Chinese and Arabic sentiment classification demonstrate that ADAN significantly outperforms state-of-the-art systems.

TACL journal version

References in corpus (3)

• Deep Generative Image Models using a Laplacian Pyramid of Adversarial Networks
• Improved Training of Wasserstein GANs
• Domain-Adversarial Neural Networks

Cited by in corpus (20)

• Adversarial Learning for Neural Dialogue Generation
• Controllable Invariance through Adversarial Feature Learning
• Cross-lingual Text-independent Speaker Verification using Unsupervised Adversarial Discriminative Domain Adaptation
• Stacked Generative Adversarial Networks
• Adversarial Training in Affective Computing and Sentiment Analysis: Recent Advances and Perspectives
• Low-Resource Text Classification using Domain-Adversarial Learning
• Cross-Lingual Sentiment Analysis Without (Good) Translation
• Improving Neural Machine Translation with Conditional Sequence Generative Adversarial Nets
• Adversarial Removal of Demographic Attributes from Text Data
• Adversarial Learning for Chinese NER from Crowd Annotations
• Cross-Lingual Transfer Learning for Question Answering
• Cross-lingual Distillation for Text Classification
• Chinese Discourse Segmentation Using Bilingual Discourse Commonality
• Adversarial Domain Adaptation for Duplicate Question Detection
• Adaptive Semi-supervised Learning for Cross-domain Sentiment Classification
• Neural Task Representations as Weak Supervision for Model Agnostic Cross-Lingual Transfer
• Adversarial Connective-exploiting Networks for Implicit Discourse Relation Classification
• One "Ruler" for All Languages: Multi-Lingual Dialogue Evaluation with Adversarial Multi-Task Learning
• LTG-Oslo Hierarchical Multi-task Network: The importance of negation for document-level sentiment in Spanish
• Addressee and Response Selection for Multilingual Conversation