Online media has become the stage for expressing interest and criticize the things and policies of the organizations and government. Every internet user has the freedom to express their view and share their felling on this platform. In the context of India, the Government promulgated the farmers related to three acts, and the population of India especially farms are opposing these three acts. The farmer community and other related communities are worried about the implementation of these acts. At least 70% of the population in India dependents on agriculture and they have shown resentment against these acts on a social media platform and have expressed their reviews. These communities have been used more than one language for expressing their views about these acts. Multiple languages have been mixed with those having different rules of grammars, which have become a challenging task for researchers to analyses the sentiments from such platforms. In this paper, the author projected statistical technique to perform sentiment analyse based on extracted agriculture tweets containing mixed content of English and Punjabi languages. In addition to this, we focused on the accuracy and performance of the agriculture data-set for the prediction of the sentiment on the tested data-set.