Sentiment Analysis Using SVM and Deep Neural Network

Abstract

With the incremental growth of Web based platforms, Automated Review sentiment analysis has become main focus, with the very primary purpose is to get the implicit information by the reviews for the better acknowledgement of the customers. Conventional methods have made tremendous growth on the basic level and are not very much reliable including the sentiment analysis, inspite the rapid growth of data scale and processing high dimension of data faster, effectively and accurately becomes the biggest drawbacks. The primary focus of the paper is the use of Support vector machines, embedding & deep neural networks which are suitable for the high dimensional data analysis, and proposes the structure of sentiment analysis using deep learning. This will act as the generalized automated review system for various industries Such as E-Commerce platform, YouTube comments, movie review rating etc. with more high precision and has potential to replace orthodox star-based rating System. Experiments which were carried on different data scales and various domains reveals that the proposed method can be used to solve the problem with better performance and higher Accuracy.