Predicting online video popularity using machine learning
In this digital world, the rapid spread of online video content has revolutionized the way information is disseminated and consumed. As the online landscape becomes increasingly crowded, both the creators and the audience need to seek effective methods to identify popular videos and predict online v...
Saved in:
Main Author: | |
---|---|
Other Authors: | |
Format: | Thesis-Master by Coursework |
Language: | English |
Published: |
Nanyang Technological University
2024
|
Subjects: | |
Online Access: | https://hdl.handle.net/10356/173006 |
Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
Institution: | Nanyang Technological University |
Language: | English |
Summary: | In this digital world, the rapid spread of online video content has revolutionized the way information is disseminated and consumed. As the online landscape becomes increasingly crowded, both the creators and the audience need to seek effective methods to identify popular videos and predict online video popularity. In this dissertation project, I proposed to use machine learning techniques to improve the methods to predict online video popularity.
YouTube is one of the largest online video websites today, so I use the YouTube dataset to predict online video popularity, considering features selection, fusion, and min-max normalization in the dataset, I make more improvements other than the state-of-the-art research. This study predicts online video popularity using different machine learning techniques such as Random Forest, Decision Tree, and XGBOOST, after comparing their prediction accuracy, standard deviation, and other parameters, I find that the XGBOOST model has the best prediction results, and I improve the Tuned XGBOOST model to get better results. Furthermore, to validate our claims, I use cross-validation methods for assessing various parameter value combinations. The results we obtained indicate that the models and techniques we propose are highly effective and capable of predicting the popularity of online videos with both precision and accuracy. |
---|