Scholarlink

In today’s media-driven world, Twitter (now known as X) stands out due to its user base and real-time interactions making it a valuable platform for analyzing public sentiment. It aims to be a tool with applications in fields like marketing, politics, and personal interests by offering insights into trends and public opinion. This study focuses on using Python and NLP to analyze sentiments in Twitter data through machine learning methods. Sentiment analysis through machine learning involves training a model to predict whether a given text carries positive, negative, or neutral sentiment. A new dataset was compiled by gathering 11,226 tweets and subjecting them to preprocessing steps. The naive Bayes, BERT, and Bi LSTM models were tested, with their performances compared. Among the three models evaluated, BERT showed the performance with an F1 score of 98.17% surpassing the other models examined.

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

An Efficient and Innovative Approach for Sentiment Analysis in Social Media Using Machine Learning Techniques

  • 会议论文

  • 发表日期：

    2025年

查看全文
  • Debajyoti Debnath,
  • SK Shamim Aktar,
  • Yagnyasenee Sen Gupta,
  • Saptarshi Chakraborty

摘要

In today’s media-driven world, Twitter (now known as X) stands out due to its user base and real-time interactions making it a valuable platform for analyzing public sentiment. It aims to be a tool with applications in fields like marketing, politics, and personal interests by offering insights into trends and public opinion. This study focuses on using Python and NLP to analyze sentiments in Twitter data through machine learning methods. Sentiment analysis through machine learning involves training a model to predict whether a given text carries positive, negative, or neutral sentiment. A new dataset was compiled by gathering 11,226 tweets and subjecting them to preprocessing steps. The naive Bayes, BERT, and Bi LSTM models were tested, with their performances compared. Among the three models evaluated, BERT showed the performance with an F1 score of 98.17% surpassing the other models examined.