Abstract:
Nowadays people of different ages are getting used to internet service in their daily life. It has
become the most valuable source of getting information, generating ideas, reviews for a product and
choosing the right one. Millions of reviews are generated on the internet about products, so it is
difficult to understand which product is actually useful. Sentiment analysis is such a research area
which can understand and extracts the opinion from the reviews. Our research report reflects an
experimental study on product review. It has been noticed that various research has been done on
sentiment analysis fields from English text or other languages such as Chinese. But detecting
sentiment from Bengali texts still needs a lot of attention. In this research, we have worked with
Bengali product review comments or sentences and get out the most popular products. We have
considered them either positive or negative. To bring a better result, we have used four methods and
found a satisfactory accuracy.