A Semi-supervised Approach for a Better Translation of Sentiment in Dialectical Arabic UGT

TL;DR

This paper presents a semi-supervised neural machine translation approach to improve sentiment translation accuracy in dialectical Arabic user-generated texts, addressing low-resource challenges and reducing critical sentiment errors.

Contribution

The study introduces a semi-supervised method leveraging monolingual and parallel data, initialized by a cross-lingual language model, to enhance dialectical Arabic translation quality.

Findings

Significant improvement in sentiment translation accuracy.

Reduction in critical sentiment polarity errors.

Effective use of semi-supervised training with limited parallel data.

Abstract

In the online world, Machine Translation (MT) systems are extensively used to translate User-Generated Text (UGT) such as reviews, tweets, and social media posts, where the main message is often the author's positive or negative attitude towards the topic of the text. However, MT systems still lack accuracy in some low-resource languages and sometimes make critical translation errors that completely flip the sentiment polarity of the target word or phrase and hence delivers a wrong affect message. This is particularly noticeable in texts that do not follow common lexico-grammatical standards such as the dialectical Arabic (DA) used on online platforms. In this research, we aim to improve the translation of sentiment in UGT written in the dialectical versions of the Arabic language to English. Given the scarcity of gold-standard parallel data for DA-EN in the UGT domain, we introduce a semi-supervised approach that exploits both monolingual and parallel data for training an NMT system initialised by a cross-lingual language model trained with supervised and unsupervised modeling objectives. We assess the accuracy of sentiment translation by our proposed system through a numerical 'sentiment-closeness' measure as well as human evaluation. We will show that our semi-supervised MT system can significantly help with correcting sentiment errors detected in the online translation of dialectical Arabic UGT.