Sentiment Analysis of company reviews using Machine Learning

Employees and job seekers today evaluate their experiences through online reviews of the company. These reviews can also be used by companies to understand their employee’s perception. This study aims to extract aspects of importance from the user reviews and the overall sentiment is classified. The suggested method entails pre-processing user reviews with a variety of methodologies, followed by sentiment score based on aspects and feature extraction. In the proposed system, sentiment is classified using Multinomial Nave Bayes (MNB), Decision Trees (DTs), Support Vector Machines (SVM), and Random Forest (RF) classifiers. Practical implementation is done on a glassdoor review dataset containing reviews from current and former employees of Google company is used. Results implicate that Random Forest produces best results as compared to the other algorithms.