Presenter: Lian Xue, Westminster College, Computer Science
Authors: Lian Xue
Faculty Advisor: Jingsai Liang, Westminster College, Computer Science
Institution: Westminster College
This research is aimed to help restaurant owners in Salt Lake City, and provide a way to get a better understanding of their customers and the well-being of their business based on reviews of their restaurants. In this research, we selected 53 different restaurants in Salt Lake City and scraped their most recent 80-100 google map reviews. With 20% data as training set, we labeled each review in this training set with different classification: positive, conflicted, neutral, and negative. We built a classification model using Bayes classifier which can automate the sentiment analysis of reviews of restaurants. We explored the benefits of using our sentiment classifying system to replace the current star review system on Google map. In addition, we discussed the correlation between time and review sentiments. This correlation can be used to determine the frequency of their reviews during the different time periods, and sentiment of reviews of a certain restaurant over different times, especially the contrast between the time before Covid-19 and after.