International Journal of Computer Applications |
Foundation of Computer Science (FCS), NY, USA |
Volume 185 - Number 18 |
Year of Publication: 2023 |
Authors: Tamim Al Mahmud, Sazeda Sultana, Antara Mondal |
10.5120/ijca2023922897 |
Tamim Al Mahmud, Sazeda Sultana, Antara Mondal . A New Technique to Classification of Bengali News Grounded on ML and DL Models. International Journal of Computer Applications. 185, 18 ( Jun 2023), 15-21. DOI=10.5120/ijca2023922897
News classification is the process of categorizing news articles into predefined categories based on their content. It involves utilizing techniques such as machine learning and NLP to analyze the textual features of news articles and assign them to relevant categories. This enables efficient organization and retrieval of news articles, aiding users in accessing information based on their interests. Many topics are covered by news, including local, national, and international news, politics, business, sports, and entertainment.In our native Bengali language, there are numerous classifications for news. The motivation for working in this research is because of the limited research in Bengali text and the scarcity of resources. This study proposed a new technique for the categorization of the news based on a comparison among different machine learning and deep learning model results. First, we have collected more than 20K data. Afterwards, applied several machine learning and deep learning models to pre-processed and cleaned data. Then apply our proposed technique on evaluated results produced by the machine learning and deep learning models to make a final decision to identify the news category.In our proposed technique, we counted the predicted categories from three types of Machine Learning models and two types of Deep Learning models. Finally, we computed the maximum predicted class and assigned the category for the specific news.