, Ābēga (Known as Emotion in Bangla), a fundamental and significant part of who we are is shaped by our emotions. Though it may not always be obvious, everything we do, and say reflects some of our emotions in some way. Human behavior can be understood if we analyze these emotions from data. Text, speech, and facial expressions are all examples of this data. The challenge of quantifying emotions is highly difficult, and although significant progress has been made, there is still much room for improvement. Since social networking websites have become so popular, many individuals have developed a strong interest in exploring the material that is posted on these different websites. By studying this data across the Internet, we are traversing the whole continent, all of its cultures and communities. The work for regional languages is limited in this particular domain. So, the purpose of this paper is to propose a model which can accurately predict or classify emotions in regional languages.