Computer Science and Engineering in Health Services. 5th EAI International Conference, COMPSE 2021, Virtual Event, July 29, 2021, Proceedings

Research Article

Sentiment Analysis Model on Twitter About Video Streaming Platforms in Mexico

  • @INPROCEEDINGS{10.1007/978-3-030-87495-7_6,
        author={Rosalia Andrade-Gonzalez and Roman Rodriguez-Aguilar},
        title={Sentiment Analysis Model on Twitter About Video Streaming Platforms in Mexico},
        proceedings={Computer Science and Engineering in Health Services. 5th EAI International Conference, COMPSE 2021, Virtual Event, July 29, 2021, Proceedings},
        proceedings_a={COMPSE},
        year={2021},
        month={9},
        keywords={Text mining Sentiment analysis Analytical intelligence Streaming Satisfaction levels},
        doi={10.1007/978-3-030-87495-7_6}
    }
    
  • Rosalia Andrade-Gonzalez
    Roman Rodriguez-Aguilar
    Year: 2021
    Sentiment Analysis Model on Twitter About Video Streaming Platforms in Mexico
    COMPSE
    Springer
    DOI: 10.1007/978-3-030-87495-7_6
Rosalia Andrade-Gonzalez1, Roman Rodriguez-Aguilar2
  • 1: Universidad Anahuac Mexico
  • 2: Universidad Panamericana

Abstract

This work addresses the analysis of the content of the comments on Twitter in the period from December 2020 to February 2021 on the video streaming platforms in Mexico: Netflix, Disney+ and Prime Video. The analysis involves the extraction of comments on Twitter, cleaning the text and the development of a supervised support model for Text Mining for the sentiment classification of tweets in the categories: Positive, Negative or Neutral (spam); as well as the use of resampling techniques to measure the variability of the model’s performance and improve the precision of its parameters. The result allows the measurement of user satisfaction levels and the detection of the most dissatisfied and liked aspects of the platforms. Finally, a business intelligence dashboard was developed in Power BI for the interactive visualization of the results under different information filters.