![]() Our database are consists of multiple tables that describe different aspects of the data which allow us to do effective joining and querying. In addition to the data processing pipeline we have developed, we also created a database to store the Yelp’s review data. We obtained our data from Yelp’s online data challenge, utilized machine learning techniques as well as natural language processing tools to retrieve insights from the data, and fit statistical models to the data so that we are able to access the most relevant keywords in the reviews that affect review scores. ![]() In the second part of this work, we examined Yelp’s merchant review data in the hope to retrieve useful information for customers to better understand a particular merchant and for merchants to improve their businesses. The true and fake reviews in the data set help us train a model that predicts if a given review is fake or not. In the first part of this work, we are examining the fake review detection data, which consists of 350,000 user reviews. Specifically, our group mainly focused on improving the customers’ understanding of merchants, market knowledge for new business owners, and existing merchants’ awareness about restaurants’ features. Utilizing our complementary plug-in prototype, yelp users, whether they are business owners or consumers, are able to find information that better meet their preferences. In order to improve Yelp users’ experience, we dived deep into Yelp’s open datasets as well as other data centers to retrieve useful information. However, Yelp only provides us a holistic view about restaurant, such as giving overall review score or ratings and only a few reviews out of thousands of reviews. Yelp is currently the most widely used restaurant and merchant information software across United States. Machine Learning and Visualization with Yelp Dataset
0 Comments
Leave a Reply. |
AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |