The digital era stages on taxation: An experimental study of text mining and pattern recognition for controlling tax on business online transaction
© 2017 Serials Publications Pvt. Ltd. The main idea of this research is to improve tax control for online businessmen who have not yet reported tax and or not yet have NPWP (Nomor Pokok Wajib Pajak), so it is necessary to clarify and re-reviewed the online business players. The purpose of this research is to design the right system or model for the detection of online business, or business transactions conducted in social media for the benefit of increasing tax revenue. Tracking and imposition tax must be worn by online business in social media, except the business into the marketplace such as, ‘Tokopedia’, ‘Buka Lapak’, ‘Lazada’. Hitherto this research is conducted there is no single research that focus on this. Based on the Directorate General of Taxation this is an important issue and should be resolved soon, due to the increasingly widespread business through social media in Indonesia last 5 years. Therefore this research needs to be done to anticipate and solve the problem of research through system design model. This detection model is carried out in two stages, the first phase focused on the classification of social media users through the text mining process. The second stage is perfecting the detection of online business through the process model of Pattern Recognition. Therefore, the Indonesia’s tax problem increase and complicated particularly in regards with the absence models for control system to increase income tax from online business tax payers, especially SMEs who the major usage social media as the way for a business transactions. This research uses experimental method. Data collection was taken from the social media users through cluster sampling technique, in this study the authors use social media Facebook as experimental data. The analysis associated with text mining is to create a variety of key words that aim to detect business transactions on Facebook. The second phase of analysis implements text mining results with data mining from the bank through design patterns based on time and nominal value. The results of this design model is expected to improve control system and tax revenue.