Identifying the Writing Style of Bangla Language Using Natural Language Processing
Syeda Farjana Shetu, Mohd. Saifuzzaman, Masuma Parvin, Nazmun Nessa Moon, Ridwanullah Yousuf, Sharmin Sultana
Abstract
Bangla is one of the 8th major spoken languages around the world and like other widely spoken languages, it is a very morphologically rich language. It has two styles, one is standard literary style, known as Sadhu Bhasha and the other one is a standard colloquial style which is known as Cholito Bhasha. Mixing both the styles in a written document is considered as a grammatical error in Bangla language known as Guruchondali Dosh. This research aims to develop an algorithm to identify the style of a Bangla paragraph i.e. whether it is in Sadhu Bhasha or Cholito Bhasha from a given Bangla paragraph input. It's a contribution towards finding the Goruchondali Dosh which is a common grammatical mistake in written Bangla language as it was observed that a number of research work for identifying Bangla grammar mistakes is not so notable whereas it is a common trend in other language researchers.