Global Signature Consultancy
Conference Paper

Data-driven Approach for Selection of an Ensemble Model of Profane Words Detection in Social Media

Abdulrehman A. Mohamed, George Okeyo, Michael Kimwele

2021 Jomo Kenyatta University of Agriculture and Technology (JKUAT), Kenya

Abstract / Description

This conference paper presents a data-driven ensemble approach for detecting profane language in social media platforms. The proposed Data-driven Ensemble Model (DEM) combines algorithm tuning, feature engineering, crowd-sourcing analytics, and ensemble learning techniques to improve profanity detection accuracy. Using Twitter datasets and WEKA for experimentation, the study demonstrated an average accuracy of 94.94%, outperforming the baseline Support Vector Machine (SVM) model, which achieved 93.33%. The research contributes to scalable and accurate automated content moderation techniques for online social platforms.

Keywords

Artificial Intelligence Machine Learning Ensemble Learning Social Media Analytics Profane Word Detection Natural Language Processing Data Mining Feature Engineering Algorithm Tuning Crowd-sourcing Analytics Twitter WEKA

Suggested Citation

Abdulrehman A. Mohamed, George Okeyo, Michael Kimwele. (2021). Data-driven Approach for Selection of an Ensemble Model of Profane Words Detection in Social Media. 2021 International Conference on Artificial Intelligence, Big Data, Computing and Data Communication Systems (icABCD).
View Publication Download PDF

Related Publications

View All Publications

Interested in Research Collaboration?

Partner with Global Signature Consultancy for research, surveys, data analytics, institutional studies and evidence-based reporting.

Contact Me