Scholarlink

The world witnessed a large-scale adaptation of internet and computer-based practices for traditional and informal education both across all levels since the year 2020. In this process, platforms such as YouTube have gained increased importance as educational tools in the everyday lives of students. It is therefore important to filter, understand and analyse the content that can impact education positively. As YouTube is an open platform, the quantity of videos catering to a certain topic can be huge, in such a case understanding the pattern of the response that viewers have to such videos may prove to be an important method to classify and determine useful good quality educational content. Our study has focused on the application of natural language processing (NLP) and sentiment analysis techniques to scrutinize user comments on the educational videos available on YouTube. Leveraging the Natural Language Toolkit (NLTK) library and Valence Aware Dictionary and Sentiment Reasoner (VADER) sentiment analyzer which examines viewer comments to understand the emotions and opinions expressed towards educational videos. Apart from helping viewers, such analysis can provide educators, content creators, and platform administrators with valuable insights, facilitating targeted enhancements in content delivery and instructional strategies to optimize audience engagement. The process includes data collection, comment extraction and sentiment analysis using of the YouTube Data API and various machine learning models such as Multinomial Naive Bayes, Logistic Regression, Support Vector Machine, Decision Tree, and Random Forest, with SVM demonstrating superior accuracy in the sentimental analysis. The study includes the development of a seamless user interface integrated with a latest processes such as a telegram bot (“Sentiment_Analysis_MD_bot”), streamlining the distribution of sentiment analysis results. Through this exploration, comprehensive insights and practical implications for harnessing sentiment analysis in improving educational content on digital platforms can be gained.

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

Understanding Viewer Sentiment on Online Educational Content: An Analysis Framework for a Video Streaming Platform Using Natural Language Processing

  • 会议论文

  • 发表日期：

    2026年

查看全文
  • Mohammad Vohra,
  • T. P. Singh,
  • Vidya Kumbhar,
  • Indraneel Krishna Kulkarni

摘要

The world witnessed a large-scale adaptation of internet and computer-based practices for traditional and informal education both across all levels since the year 2020. In this process, platforms such as YouTube have gained increased importance as educational tools in the everyday lives of students. It is therefore important to filter, understand and analyse the content that can impact education positively. As YouTube is an open platform, the quantity of videos catering to a certain topic can be huge, in such a case understanding the pattern of the response that viewers have to such videos may prove to be an important method to classify and determine useful good quality educational content. Our study has focused on the application of natural language processing (NLP) and sentiment analysis techniques to scrutinize user comments on the educational videos available on YouTube. Leveraging the Natural Language Toolkit (NLTK) library and Valence Aware Dictionary and Sentiment Reasoner (VADER) sentiment analyzer which examines viewer comments to understand the emotions and opinions expressed towards educational videos. Apart from helping viewers, such analysis can provide educators, content creators, and platform administrators with valuable insights, facilitating targeted enhancements in content delivery and instructional strategies to optimize audience engagement. The process includes data collection, comment extraction and sentiment analysis using of the YouTube Data API and various machine learning models such as Multinomial Naive Bayes, Logistic Regression, Support Vector Machine, Decision Tree, and Random Forest, with SVM demonstrating superior accuracy in the sentimental analysis. The study includes the development of a seamless user interface integrated with a latest processes such as a telegram bot (“Sentiment_Analysis_MD_bot”), streamlining the distribution of sentiment analysis results. Through this exploration, comprehensive insights and practical implications for harnessing sentiment analysis in improving educational content on digital platforms can be gained.