A Review of Multilingual Sentiment Analysis

Multilingual sentiment analysis evaluates people’s opinions and attitudes in a scenario of multiple languages. In today’s digitally globalized world, online platforms serve as spaces where people from different cultural and linguistic backgrounds share their opinions, producing a significant volume of multilingual data every day. Such multilingual data holds substantial value, highlighting the growing importance of multilingual sentiment analysis. This paper reviews the development of multilingual sentiment analysis over the past decade, covering preprocessing methods, classification methods, feature extraction techniques, evaluation metrics, representative datasets, and experimental results. Additionally, this paper discusses the limitations of current research and future prospects of multilingual sentiment analysis.