YouTube Spam Comment Detection System

EasyChair Preprint 15711

Abstract

Spamming involves posting irrelevant and unsolicited comments on social media platforms or video-sharing sites.. Bots post these messages to reduce ranking or disturb users' viewing experience which ultimately reduces the rank of the video or the post. Besides that, spam comments can also include malicious links that can steal user's sensitive data when clicked. Spamming, often manual, is prevalent on competitive platforms. While data mining can mitigate some forms of spam, this project automates spam comment detection on YouTube using machine learning techniques. We'll leverage a dataset of YouTube spam comments from the UCI Machine Learning Repository and apply the Count-Vectorizer and Support Vector Machine algorithm for clustering on the given dataset using python programming.

Keyphrases: Multi Layer Perceptron, Spam Detection, Supervised Machine Learning, spam filtering, text analysis

@booklet{EasyChair:15711,
  author    = {Siladitya Mukherjee and Soumya Dey and Anal Acharya},
  title     = {YouTube Spam Comment Detection System},
  howpublished = {EasyChair Preprint 15711},
  year      = {EasyChair, 2025}}