Related Experiment Video
Updated: Jun 28, 2026
08:53
Integrating Computerized Linguistic and Social Network Analyses to Capture Addiction Recovery Capital in an Online Community
Published on: May 31, 2019
Machine learning and deep learning-based approach to categorize Bengali comments on social networks using fused
Khandaker Mohammad Mohi Uddin1, Hasibul Hamim2, Mst Nishat Tasnim Mim2
1Department of Computer Science and Engineering, Southeast University, Dhaka, Bangladesh.
Plos One
|October 3, 2024
Summary
This study developed advanced machine learning and deep learning models to detect online harassment in Bengali comments. A hybrid model achieved 99.34% accuracy, significantly improving online safety and psychological well-being.
Area of Science:
- Computational Linguistics
- Artificial Intelligence
- Social Computing
Background:
- Online harassment poses significant threats to psychological wellness and academic success.
- Early detection of online harassment is crucial for mitigating its negative consequences.
- Social media platforms necessitate robust tools for identifying and preventing cyberbullying.
Purpose of the Study:
- To develop and evaluate machine learning (ML) and deep learning (DL) models for detecting online harassment in Bengali comments.
- To create a criticism-free online environment by identifying and eliminating cyberbullying.
- To improve the accuracy and efficiency of online harassment detection systems.
Main Methods:
- Utilized natural language processing (NLP) techniques, including tokenization and padding.
- Applied term frequency-inverse document frequency (TF-IDF) with count vectorizer for feature extraction.
- Implemented various ML models (MLP, K-NN, XGBoost, etc.) and DL models (DNN, CNN, C-LSTM, BiLSTM).
- Combined two datasets to create a corpus of 94,000 Bengali comments for training and validation.
Main Results:
- A hybrid ML model (MLP+SGD+LR) demonstrated superior performance.
- The hybrid model achieved 99.34% accuracy, 99.34% precision, 99.33% recall, and 99.34% F1 score for multi-label classification.
- The binary classification model achieved an accuracy of 99.41%.
Conclusions:
- The proposed hybrid ML model is highly effective for detecting online harassment in Bengali text.
- Advanced ML and DL techniques can significantly enhance online safety and user well-being.
- The findings provide a strong foundation for developing real-time cyberbullying detection systems.
Related Concept Videos
01:26Classification of Systems-I
Linearity is a system property characterized by a direct input-output relationship, combining homogeneity and additivity.
Homogeneity dictates that if an input x(t) is multiplied by a constant c, the output y(t) is multiplied by the same constant. Mathematically, this is expressed as:
Homogeneity dictates that if an input x(t) is multiplied by a constant c, the output y(t) is multiplied by the same constant. Mathematically, this is expressed as:
01:31Classification of Systems-II
Continuous-time systems have continuous input and output signals, with time measured continuously. These systems are generally defined by differential or algebraic equations. For instance, in an RC circuit, the relationship between input and output voltage is expressed through a differential equation derived from Ohm's law and the capacitor relation,
01:29Aggregates Classification
Aggregate classification is generally based on its size, petrographic characteristics, weight, and source. Size classification ranges from coarse to fine aggregates, defined by the size of the particles. Coarse aggregates are particles that do not pass through ASTM sieve No. 4, and aggregates that pass through the sieve are fine aggregates.
Petrographic classification groups aggregates based on common mineralogical characteristics. Some of the common mineral groups found in aggregates are...
Petrographic classification groups aggregates based on common mineralogical characteristics. Some of the common mineral groups found in aggregates are...