• Medientyp: E-Artikel
  • Titel: Multi-Class Classification of the YouTube Comments using Machine Learning
  • Beteiligte: Nawaz, Shoaib; Rizwan, Muhammad; Yasin, Samina; Ahmed, Mehtab; Farooq, Umar
  • Erschienen: The University of Lahore, 2022
  • Erschienen in: Pakistan Journal of Engineering and Technology
  • Sprache: Nicht zu entscheiden
  • DOI: 10.51846/vol3iss2pp183-188
  • ISSN: 2664-2050; 2664-2042
  • Schlagwörter: General Medicine
  • Entstehung:
  • Anmerkungen:
  • Beschreibung: <jats:p>Due to huge data on Social media the people face difficulty to finding in qualitative content of the video if they find the qualitative content as per their judgment and knowledge, they do not confirm the actual content quality. People put their idea and subscriber watch the videos and put their feedback in the comments. Our purposed study help the subscribers to find the best-experienced idea based on the people personal experience here we classify the collection of comments collected using Google API’s(“Google API YouTube” n.d.) and annotate them in different classes which are Experience-positive, Experience-negative, Warning, Suggestions, Questions, Praise, these classes helps us to find the qualitative analysis of the comments but based on the these required classes. Our proposed model shows experiments that the best classifier for us is the SVM have accuracy is 89.18% and F1 score is 0.89 this shows that our model is productive and efficient for classification to help out the user as well as the author to find the best qualitative video content which is experienced and confirmed by the user’s experience(Abbas 2017).</jats:p>
  • Zugangsstatus: Freier Zugang