INTERNATIONAL JOURNAL OF SCIENTIFIC DEVELOPMENT AND RESEARCH International Peer Reviewed & Refereed Journals, Open Access Journal ISSN Approved Journal No: 2455-2631 | Impact factor: 8.15 | ESTD Year: 2016
open access , Peer-reviewed, and Refereed Journals, Impact factor 8.15
The profitability of Google's newly launched video distribution platform, YouTube, has attracted a growing user base. However, alongside its success, there has been a surge in malicious activities aimed at self-promotion or spreading viruses and malware. Due to YouTube's limited comment moderation tools, spam levels have risen dramatically, prompting owners of popular channels to disable the comments section. Filtering automatic comment spam on YouTube poses a challenge for conventional classification methods, given the brevity of messages and their frequent use of slang, symbols, and abbreviations. In this study, we evaluated several high-performance classification techniques for this purpose. Statistical analysis of the results indicates, with a confidence level of 99.9%, that decision trees, logistic regression, Bernoulli Naïve Bayes, random forests, linear and Gaussian SVMs exhibit statistically equivalent performance. Based on these findings, we propose TubeSpam, an accurate online system for filtering comments posted on YouTube, aiming to mitigate the proliferation of spam while ensuring a plagiarism-free approach.
Keywords:
You tube spam ,comments detection, Support Vector Machine.
Cite Article:
"IMPROVING ACCURACY IN CLASSIFICATION OF YOUTUBE SPAM COMMENT DETECTION", International Journal of Science & Engineering Development Research (www.ijsdr.org), ISSN:2455-2631, Vol.9, Issue 4, page no.703 - 707, April-2024, Available :http://www.ijsdr.org/papers/IJSDR2404098.pdf
Downloads:
000338171
Publication Details:
Published Paper ID: IJSDR2404098
Registration ID:210820
Published In: Volume 9 Issue 4, April-2024
DOI (Digital Object Identifier):
Page No: 703 - 707
Publisher: IJSDR | www.ijsdr.org
ISSN Number: 2455-2631
Facebook Twitter Instagram LinkedIn