Sentiment Analysis of Reviews

Date
2021-01-01
Authors
Chitla, Pravalika Ravikumar
Major Professor
Anthony Townsend
Advisor
Committee Member
Journal Title
Journal ISSN
Volume Title
Publisher
Altmetrics
Authors
Research Projects
Organizational Units
Journal Issue
Series
Department
Information Systems and Business Analytics
Abstract

Abstract

The process of understanding a given text's emotion that one can use to analyze the product/service reviews is called sentiment analysis. Furthermore, classify them as either Positive, Neutral, or Negative sentiments. Learning the human masses' sentiments towards distinct entities and products facilitates better contextual advertisements, recommendation systems, and market trends analysis. Sentiment Analytics extract sense from human speech and have the power to suggest business insights into large amounts of data such as the opinions of customers, insights that inspire business decisions. Sentiment analysis systems allow companies to make sense of the sea of unstructured text by automating business processes, getting actionable insights, and saving hours of manual data processing by making businesses more efficient. I have chosen the Yelp dataset focusing on reviews of restaurants in Illinois to determine whether they are positive or negative sentiments and develop a model for their classification. This research project concentrates on predicting whether customers have a positive or negative attitude towards a restaurant. I have built a Machine learning model LSTM and used other classifier models such as Naïve Bayes, Bernoulli's, Logistic Regression, Stochastic Gradient Descent, and Support Vector Machine to train the models. I have used evaluation matrices such as prediction, accuracy, recall, and F_1 score to evaluate the classifier's performance on the test sample. Based on the evaluation matrix results, the goal is to find trends in how positive and negative reviews are written and come up with a model for their classification. The results show that SVM is the best model for classifying the reviews to either positive or negative as it gives the highest precision compared to the other classification models.

Comments
Description
Keywords
Citation
DOI
Source