Abstract: Social media platforms have become a crucial part of our daily lives and play significant roles in various aspects such as business, entertainment, marketing, education, media, and communication. Among these platforms, YouTube has gained massive popularity as the most widely used platform for sharing videos due to its unique behavior. The platform allows anyone to create an account and upload videos of their choice, which can be viewed by millions of people. This has become a trend in the entertainment industry, making it easy for videos to reach users and gain popularity online. However, not all YouTube videos become popular, and many channel owners take various actions to make their videos popular.

This research study uses sentiment analysis and feature extraction methods to derive the set of features required to consider in the development of YouTube videos. By analyzing user comments, the study aims to discover the most important trending videos related to user video types and the most trending videos that users will want to see.

The study uses machine learning methods to analyze the trending features and identify key recommendations for the users. The results of the study will enable YouTubers to create videos that resonate with their audience and increase their chances of being popular.

The study on YouTube trending videos and Support Vector Machine (SVM) algorithm has revealed the importance of views, likes, and dislikes in determining a video's trend. The SVM algorithm uses these factors to identify and predict which videos will become popular on the platform. This research study provides recommendations to YouTubers on how to create videos that can become trending by analyzing user comments, identifying their requirements, and using machine learning methods to derive the necessary features.

Keywords: KNN (K-Nearest Neighbors), SVM (Support Vector Machine), HDFS (Hadoop Distributed File System), CNN (Convolutional Neural Network), Feature Extraction methods, Machine Learning, MLP (Multi-Layer Perception), popularity

PDF | DOI: 10.17148/IJARCCE.2023.124211

Open chat
Chat with IJARCCE