Improved Sentiment Detection via Label Transfer from Monolingual to Synthetic Code-Switched Text
This addresses the lack of labeled data for code-switched text, particularly for minority languages, enabling better sentiment and hate speech detection for multilingual speakers.
The paper tackled the problem of poor sentiment detection on code-switched text by synthesizing labeled code-switched text from monolingual data, achieving accuracy improvements of 1.5% to 7.20% across language pairs and up to 6% gains in hate speech detection.
Multilingual writers and speakers often alternate between two languages in a single discourse, a practice called "code-switching". Existing sentiment detection methods are usually trained on sentiment-labeled monolingual text. Manually labeled code-switched text, especially involving minority languages, is extremely rare. Consequently, the best monolingual methods perform relatively poorly on code-switched text. We present an effective technique for synthesizing labeled code-switched text from labeled monolingual text, which is more readily available. The idea is to replace carefully selected subtrees of constituency parses of sentences in the resource-rich language with suitable token spans selected from automatic translations to the resource-poor language. By augmenting scarce human-labeled code-switched text with plentiful synthetic code-switched text, we achieve significant improvements in sentiment labeling accuracy (1.5%, 5.11%, 7.20%) for three different language pairs (English-Hindi, English-Spanish and English-Bengali). We also get significant gains for hate speech detection: 4% improvement using only synthetic text and 6% if augmented with real text.