Scholarlink

Twitter is very important for measuring how people feel about airline companies. This has made scientists more interested in using social media to understand what passengers think and experience. Businesses are using Twitter to get fast opinions from customers because old ways of knowing what they think were slow. We studied a plane data set from Kaggle to see how machine learning methods (Naive Bayes, SVM and others) could be used for finding out if people feel good or bad about things. We also looked at different ways to choose the best features using feature selection methods that can help improve accuracy. Looking at several types of classifiers with and without choosing key features shows how Twitter data helps make sentiment analysis better in the airline world.

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

Optimizing Sentiment Analysis in Airline Feedback: Feature Selection Impact on Machine Learning Models

  • 会议论文

  • 发表日期：

    2025年

查看全文
  • Arpan Soni,
  • Brigunandan Sharma,
  • Snigdha Bangal

摘要

Twitter is very important for measuring how people feel about airline companies. This has made scientists more interested in using social media to understand what passengers think and experience. Businesses are using Twitter to get fast opinions from customers because old ways of knowing what they think were slow. We studied a plane data set from Kaggle to see how machine learning methods (Naive Bayes, SVM and others) could be used for finding out if people feel good or bad about things. We also looked at different ways to choose the best features using feature selection methods that can help improve accuracy. Looking at several types of classifiers with and without choosing key features shows how Twitter data helps make sentiment analysis better in the airline world.