Scholarlink

In this comprehensive research, we explore multilingual sentiment analysis, aiming to classify sentiments as positive, negative, or neutral across languages. Our dataset includes 192,839 examples in English, Japanese, German, French, Spanish, Chinese, Indonesian, and Hindi. We assess three models: XGBoost (62% accuracy), SVM (64.2% accuracy), and the multilingual XLM-RoBERTa-base, which excels with an impressive 78.37% accuracy. XLM-RoBERTa-base’s adaptability to diverse languages makes it a crucial asset in multilingual sentiment analysis.

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Advancing Multilingual Sentiment Understanding with XGBoost, SVM, and XLM-RoBERTa

  • 会议论文

  • 发表日期：

    2025年

查看全文
  • Arya Gaikwad,
  • Pranav Belhekar,
  • Vinayak Kottawar

摘要

In this comprehensive research, we explore multilingual sentiment analysis, aiming to classify sentiments as positive, negative, or neutral across languages. Our dataset includes 192,839 examples in English, Japanese, German, French, Spanish, Chinese, Indonesian, and Hindi. We assess three models: XGBoost (62% accuracy), SVM (64.2% accuracy), and the multilingual XLM-RoBERTa-base, which excels with an impressive 78.37% accuracy. XLM-RoBERTa-base’s adaptability to diverse languages makes it a crucial asset in multilingual sentiment analysis.