Sentiment Analysis for Bangla Sentences using Convolutional Neural Network

Sentiment analysis, also known as opinion mining or emotion analysis, is a process to determine emotional reaction of people towards an interaction or an event. An opinion may be positive, negative or neutral depends on individuals judgment or evaluation towards a topic or an event. Usually sentiments can be varied in cultures and languages. On sentiment analysis, an extensive amount of research works can be seen for well-resourced languages like English, Japanese etc. However, such works are relatively less observed for low-resourced language like Bangla. In this paper, we propose a framework that analyzes sentiments from texts written in Bangla. In our proposal, we use Bangla comments and generate a classification model. The model is generated by a neural network variance called Convolutional Neural Network. The classifier model obtains a classification accuracy of 99.87%, which is 6.87% better than the available state-of-the art Bangla sentiment classifier.