DoR

Scopus Indexed Publications

Title: Neural Network Based Bengali News Headline Multi Classification System: Selection of Features describes Comparative Performance

Author: Sharun Akter Khushbu, Abu Kaisar Mohammad Masum, Sheikh Abujar, Syed Akhter Hossain,

Abstract: The modern era is gradually developing in all sectors. Moreover, development is necessary to research. Accordingly natural language processing is the expanding area of process text. At present year, memorizing the data is very tough due to the rapidly growing volume of data. Newspapers are a great habit to the whole world of all ages. To acquire a variety of knowledge from different segments is entertaining by themselves. According to this work using bengali news headline may be more specific to others by defining its news type. Therefore the machine can smartly review the sequence of sentences within reach output to find newstype. With experiment, we connect our approach by Neural Network of adoption with 90% accuracy performance. Coming with a momentous outcome we've done Multi Classification reached at SVM, NB, Logistic Regression, Neural Network, Random forest applying bengali dataset.

Keywords: Natural Language Processing , process text , Newspapers , bengali news headline , SVM , NB , Logistic Regression , Neural Network , Random forest

Journal or Conference Name: 11th International Conference on Computing, Communication and Networking Technologies, ICCCNT 2020