Multilingual Sentiment Analysis

Project Overview

Extended sentiment classification work beyond Arabic to a multilingual setting, evaluating how well transformer-based approaches generalize across languages with different morphological structures and resource levels.

Technical Details

• Architectures: Multilingual transformer models (e.g. mBERT, XLM-R) evaluated against language-specific fine-tuned models.
• Languages: [List the actual languages covered, e.g. Arabic, French, English.]
• Approach: Compared zero-shot cross-lingual transfer against per-language fine-tuning, to assess how much language-specific adaptation is actually needed.

Challenges

Cross-lingual transfer tends to work well between related, well-resourced languages but degrades for morphologically distant or low-resource ones, echoing the same pattern seen in the Arabic-specific sentiment work. Part of the evaluation was quantifying exactly where and how much that gap appears.

Outcome

[Status: in progress / completed — update this line.] [Add your actual finding, e.g. “Zero-shot transfer retained X% of in-language performance for related languages, but dropped to Y% for Arabic without fine-tuning.”]

My Role

Designed the cross-lingual evaluation protocol and ran the comparative experiments across languages and model variants.