MAGE: Multi-Head Attention Guided Embeddings for Low Resource Sentiment Classification

TL;DR

This paper presents MAGE, a novel model combining multi-head attention guided embeddings and language-independent data augmentation to improve sentiment classification in low-resource Bantu languages.

Contribution

Introduction of MAGE, a model integrating multi-head attention and data augmentation to enhance low-resource language text classification.

Findings

Improved classification accuracy on Bantu languages

Effective data augmentation across linguistic contexts

Robust handling of syntactic and semantic features

Abstract

Due to the lack of quality data for low-resource Bantu languages, significant challenges are presented in text classification and other practical implementations. In this paper, we introduce an advanced model combining Language-Independent Data Augmentation (LiDA) with Multi-Head Attention based weighted embeddings to selectively enhance critical data points and improve text classification performance. This integration allows us to create robust data augmentation strategies that are effective across various linguistic contexts, ensuring that our model can handle the unique syntactic and semantic features of Bantu languages. This approach not only addresses the data scarcity issue but also sets a foundation for future research in low-resource language processing and classification tasks.