Text Mining and Sentiment Analysis of Newspaper Headlines

Text analytics are well-known in the modern era for extracting information and patterns from text. However, no study has attempted to illustrate pattern priorities of newspaper headlines Bangladesh using a combination text techniques. The purpose this paper is examine words that appeared on front page daily English Bangladesh, Daily Star, 2018 2019. elucidation era’s possible social political context was also word patterns. employs three widely used contemporary mining techniques: clouds, sentiment analysis, cluster analysis. cloud reveals election, kill, cricket, Rohingya-related terms more than 60 times 2018, whereas BNP, poll, AL, Khaleda 80 These indicated country’s passion turmoil, issues. Furthermore, analysis fear negative emotions 600 times, anger, anticipation, sadness, trust, positive-type came up 400 both years. Finally, clustering method demonstrates politics, deaths, digital security act, Rohingya, cricket-related exhibit similarity belong similar group 2019, rape, road, fire-related clustered alongside similar-appearing group. In general, how vividly approach depicts Bangladesh’s social, political, law-and-order situation, particularly during election season cricket craze, validates significance understanding overall view country particular time an efficient manner.