News production and its spreading have rapidly been changed by the social networking sites for instance Facebook. All the Bangla news exists on social media is in textual format which is unstructured as well. Different techniques of Text mining play a vital role in order to convert those Bangla unstructured news into formative knowledge. As there are lack of analysis regarding Bangla news of Facebook posts have been introduced, present study looks for drawing a pattern that refers a constructive knowledge from huge amount of data. To accomplish that, three newspapers have been chosen, namely ProthomAlo, Juganthor and Daily NayaDiganta. Facepager tool has been used to extract data from the Facebook pages of aforementioned newspapers and later data was processed through Spyder, an environment to run Python program. Consequence stated that ?Bangladesh National Election? along with the ?Political Issues? received maximum coverage followed by the recent ?Football World Cup?. Besides, the most frequent newspaper that shares posts on Facebook is Juganthor followed by Daily Naya Diganta and ProthomAlo, respectively. It is also to be said that there is a significant resemblance between Juganthor and Daily Naya Diganta in posting identical posts on Facebook

word cloud

trending news

News similarity

text mining