Sentiment Analysis of User Reviews for the LinkedIn Application Using Support Vector Machine and Naïve Bayes Algorithm

Abstract

Social Networking Sites (SNS) have become integral communication platforms for knowledge sharing and professional connections. LinkedIn, a leading professional network, is widely utilized in today's digital era, primarily by professionals and the business community. This research focuses on analyzing user sentiment on LinkedIn through the application of the Support Vector Machine (SVM) and Naive Bayes methods. Understanding user opinions and satisfaction is important, and sentiment analysis serves as a key tool for this purpose. This study is a comparative analysis of Support Vector Machine (SVM) and Naïve Bayes algorithm for classifying user reviews of the LinkedIn application. Drawing on data from Google Play reviews, this research explores a range of user sentiment towards the LinkedIn platform, including positive, negative and neutral reviews. The application of SVM and Naive Bayes algorithms successfully classifies reviews into relevant sentiment categories. Analyzing 2000 review datasets with an 80% training and 20% testing data split, Support Vector Machines demonstrate an 80% accuracy rate, while Naïve Bayes achieves a 70% accuracy rate. The Support Vector Machines (SVM) algorithm has better accuracy than the Naïve Bayes algorithm based on the test scenarios that have been carried out.