International Journal on Advanced Science, Engineering and Information Technology, Vol. 12 (2022) No. 5, pages: 1944-1949, DOI:10.18517/ijaseit.12.5.14387

Classification of Indonesian Population's Level Happiness on Twitter Data Using N-Gram, Naïve Bayes, and Big Data Technology

I Nyoman Krisna Bayu, I Made Agus Dwi Suarjaya, Putu Wira Buana


The level of happiness is one factor that influences social interaction in the community. Therefore, the population's happiness level within the current year has become an exciting concern to be studied. Since last year, the world has been facing a COVID-19 pandemic. COVID-19 pandemic dramatically affects the happiness level of the population from a social, economic, health, education, and tourism perspective. The various affected sectors cause different levels of emotional happiness in the community in terms of social interactions in opinions and issues on social media. In addition, the number of issues on social media induce a vast data warehouse and high complexity. Big Data is a science that handles large amounts of data, which is unmanageable using traditional data processing methods or techniques. Various companies, organizations, researchers, and academics practice Big Data to extract and analyze the necessary information. Big Data is a general term used for all data collection forms of vast and complex nature. The utilization of Big Data can be valuable for a better decision-making process. This study uses Big Data Technology to evaluate the Indonesian population's happiness level on Twitter data. Method classified and technique using the N-Gram, Naïve Bayes, and Laplacian Smoothing Technique. The emotion in this research is classified into two aspects: happy and unhappy emotions. A total of 4.306.581 tweet data is classified; the obtained results revealed 39,4% happy emotion and 60,6% unhappy emotion.


Happiness level; tweet; big data; n-gram; naïve bayes; laplacian smoothing.

Viewed: 1183 times (since abstract online)

cite this paper     download