Model in Distress: Sentiment Analysis on French Synthetic Social Media

2026-04-20

AI Summary

This work addresses three key challenges in automated analysis of social media customer feedback: high annotation costs, scarcity of multilingual evaluation benchmarks, and privacy concerns. To overcome these, the authors propose a generalizable synthetic data generation approach that combines fine-tuned back-translation models with synthetic reasoning traces to produce 1.7 million French synthetic tweets from a small seed corpus. Using this data, they train a 600-million-parameter bilingual English–French language model. This study is the first to integrate synthetic reasoning traces with back-translation for sentiment analysis, substantially reducing reliance on human annotations while preserving user privacy and enabling effective cross-lingual transfer. On a human-annotated French evaluation set, the model achieves 77–79% accuracy, matching or surpassing state-of-the-art proprietary large language models and specialized encoders.

Technology Category

• Natural Language Processing: Code Generation / Program Synthesis from Natural Language
• Machine Learning: Large Multimodal Models (LMMs)
• Application Domains: Humanities & Computational Social Science

Application Category

• Social Networks and Social Media: Generative AI / large language models and their impact on social systems
• Search and Retrieval-Augmented AI: Multilingual and cross-lingual Web search
• Web Mining and Content Analysis: Large pretrained models with web data

Abstract

Automated analysis of customer feedback on social media is hindered by three challenges: the high cost of annotated training data, the scarcity of evaluation sets, especially in multilingual settings, and privacy concerns that prevent data sharing and reproducibility. We address these issues by developing a generalizable synthetic data generation pipeline applied to a case study on customer distress detection in French public transportation. Our approach utilizes backtranslation with fine-tuned models to generate 1.7 million synthetic tweets from a small seed corpus, complemented by synthetic reasoning traces. We train 600M-parameter reasoners with English and French reasoning that achieve 77-79% accuracy on human-annotated evaluation data, matching or exceeding SOTA proprietary LLMs and specialized encoders. Beyond reducing annotation costs, our pipeline preserves privacy by eliminating the exposure of sensitive user data. Our methodology can be adopted for other use cases and languages.

Problem

• sentiment analysis
• synthetic data
• multilingual
• privacy
• social media

Innovation

• synthetic data generation
• backtranslation
• reasoning traces
• multilingual sentiment analysis
• privacy-preserving NLP