Abstract:
In recent years it is noticeable that sharing text reviews on various businesses specially restaurants
through website and social media is a very common phenomenon. Online reviews reflect user’s
opinion. This huge collection of user data in terms of text reviews can be analyzed to identify
user’s sentiment and their demand also. Here users are the primary sources. Text reviews are the
complete reflection of user’s sentiment and also owned by them. Measuring user’s sentiment will
also be able to find out the market position of a Transportation system. By making the machine
learned about the total reviews, it will be able to categorize the unknown text. We collect the
necessary data for our research work from a verified source and Google Play store. We took a step
forward by combining user review texts which were collected from that website to build a model
that can predict a review asserting good or bad and Average. Key benefit of our approach is that,
by using our proposed model transport system Owners can identify the main focused term from
the review of customers and also can take future step to work on that.
We are also able to publish the position of a System by counting that how many reviews are Good
, Bad, and Average comparative to with each. As this model is based on text document, it will be
very perfect work in all terms and condition. Because text document shows almost the best
predicting result of user’s sentiment than that of star rating does.
