Abstract
Everyday millions of customers use yelp app for various purposes. Yelp app is used to find businesses nearby your location, such as ATM machines, banks, restaurants, saloons, stores etc. When you search for a specific business on Yelp, the details of the business is displayed which contain business hours, the distance between your location and the store, reviews, menus or services they provide. Sometimes there would be too many reviews written by customers and people may feel it difficult to digest them. If the business is being rated, then it would be really easy for the customers to decide whether to choose that business or not. Natural language processing technique could be used to take in various reviews written by the customers and integrate them to extract useful ratings which can be efficient and accurate. This project involves sentimental analysis of reviews provided by the customers and the dataset is provided by Yelp. The purpose of this project is to provide customers with appropriate ratings of various businesses just based on the previous reviews written by customers. Yelp dataset contains individual json files related to business hours, users, check in, reviews and tips. All these json files are inter- linked. Sentimental Analysis is performed on the reviews given by customers and related ratings are determined. This project includes data collection, data loading, and data cleaning, pre-processing and sentimental analysis of the reviews. To process Big Data, Hadoop is being used in this project. The data is loaded into Hive and processed. After the data is processed, reviews are analyzed based on sentimental analysis and related ratings are generated.