Brief state of the art in social information mining: Practical application in analysis of trends in French legislative 2024

The analysis of social media information has undergone significant evolution in the last decade due to advancements in artificial intelligence (AI) and machine learning (ML). This paper provides an overview of the state-of-the-art techniques in social media mining, with a practical application in analyzing trends in the 2024 French legislative elections. We leverage natural language processing (NLP) tools to gauge public opinion by extracting and analyzing comments and reactions from the AgoraVox platform. The study reveals that the National Rally party, led by Marine Le Pen, maintains a high level of engagement on social media, outperforming traditional parties. This trend is corroborated by user interactions, indicating a strong digital presence. The results highlight the utility of advanced AI models, such as transformers and large language models (LLMs), in capturing nuanced public sentiments and predicting political leanings, demonstrating their potential in real-time reputation management and crisis response.