Mining Better Advertisement Tool for Government Schemes Using Machine Learning
1E. Prabhakar, V.S. Suresh Kumar, Dr.S. Nandagopal and C.R. Dhivyaa
Public opinion impacts most government policies and sets the limits within which policymakers must operate. It also sets the pace of reform. Central and state governments allocate massive budget for advertising schemes. Advertisement is necessary for reaching out to maximum people. But more money is invested in the advertisement. This invested amount will be reduced if governments select proper advertisement medium (Printed/Television/Social). India is diverse behavior nation. Citizens of India have different kinds of diverse educational background, culture, etc., Habitation of people in rural, urban and tribal varies considerably. The same advertisement medium is not suitable to reach out to these diverse natured set of peoples. So this paper focuses on public opinion mining to find the best advertising medium for government schemes. Data set is collected through online. Collected citizen data include gender, qualification, government school/private school, the field of work, occupation, year of experience, city/village, age, government job/private job, type of advertising medium people are comfortable. The proposed system introduces two new classification algorithms to predict the best advertisement medium to advertise government schemes. The final results show that the performance of the proposed algorithms outperformed existing algorithms.
Government Schemes, Advertisement Tool, Classification, Class Imbalance, Mean Error Based Ensemble