Measuring the effectiveness of Government Schemes using Machine learning algorithms

Communication is being an important aspect of life for exchanging the information. Digital communication has got the real boom in the recent past. Internet has made the biggest impact on the digital communication. Social media have contributed in a big way for the communication such as facebook, gmail, Twitter, yahoo, linkedin etc. The data generated through these social media is really huge and unstructured. The processing and analyzing of this huge data is very much essential. The knowledge extraction from the analysis of this huge data could be helpful in decision making process in various domains. The paper focuses on analyzing the twitter data about various government schemes such as "swatch Bharat Abhiyan", "digital India" and "demonetization" using Naïve Bayes and Maximum entropy algorithms. At the end the effectiveness of these algorithms is compared based on their performance. Also the popularity of these schemes is analyzed using people's opinion.