Using Machine Learning models to detect the category of Bangla news articles by analyzing the key features for each of the defined categories. This project was done as part of my Final Year Thesis (CSE-400 and CSE-408).
Main project file is the following file: 5_cat_small_mlp.py
We publised two papers from this project. One of our works was published in "2017 20th International Conference of Computer and Information Technology (ICCIT), Dhaka, 2017". Here we discussed about the effects of different kernels of SVM in Bangla News Categorization. The paper can be found in the folling link:
DOI: 10.1109/ICBSLP.2018.8554844
Please cite our work if you find it useful:
Mahmud, Q. I., Chowdhury, N. I. and Masum, M. (2018). Reducing Feature Space and Analyzing Effects of Using Non Linear Kernels in SVM for Bangla News Categorization. 2018 International Conference on Bangla Speech and Language Processing (ICBSLP), Sylhet, 2018, pp. 1-6. DOI: 10.1109/ICBSLP.2018.8554844
The second paper was accepted at the "Journal of Computer Science" (Volume 16, No. 3, 2020, Pages: 378-390). You may read the paper to understand how we used category specific feature extraction technique to solve the problem of Bangla News Categorization along with a Multi Layer Perceptron model. The paper can be found at the following link:
DOI: https://doi.org/10.3844/jcssp.2020.378.390
To cite our work use the following format:
Mahmud, Q. I., Chowdhury, N. I. & Masum, M. (2020). A Multi Layer Perceptron Along with Memory Efficient Feature Extraction Approach for Bengali Document Categorization. Journal of Computer Science, 16(3), 378-390. DOI: https://doi.org/10.3844/jcssp.2020.378.390