Development of a Web-Based Automatic Sentiment Analysis Application using Support Vector Machine (SVM) Model

Abstract

This research aims to develop a web-based automatic sentiment analysis application using the Support Vector Machine (SVM) model. Through this application, users can easily analyze the sentiment of the text they input through a user-friendly interface. In the initial stage of the research, we conducted a review of various existing techniques for automatic sentiment analysis. From the review, we selected the SVM model as the main model in our application due to its effectiveness in sentiment classification. We used the Streamlit web framework to build a responsive and user-friendly user interface.

The methods we applied include data preprocessing and processing, feature extraction using TfidfVectorizer, and training the SVM model. We involved a labeled dataset to train the model and performed performance evaluation using separate test data. Our evaluation results showed that the implementation of the SVM model in our application provided excellent results in sentiment analysis with an accuracy rate of 94% using 830 data points. Our application interface is designed to be simple yet informative, allowing users to input text and quickly view sentiment analysis results.

In conclusion, we propose that the use of the SVM model in web-based automatic sentiment analysis applications can make a significant contribution to natural language processing. The application we have developed has the potential to be used in various fields such as social media monitoring, product review analysis, and understanding user opinions in general.