A Survey on Sentimental Analysis of Online Product Reviews for Various E-Shopping Websites
Punitha Devi C, Vijayalalitha, Raja Rajeswari S, Lendaa · Zenodo (CERN European Organization for Nuclear Research) · 2018
The rapid growth of internet is a wide source of learning and decision making. Thousands of reviews are generated each day for the online products. Analyses of these opinions are very important because it is helpful in determining the users' perspective about the product and also for making business decisions in the organizations. The sentiment analysis system is build on the basis of identifying the words and classifying its sentiments from the e-shopping websites using data mining approach. The proposed system selects the words from the learning dataset and then classifies them into objects and sentiments based on a feature extraction methodology. A Quality Data miner technique has been adapted to study about the polarity of the each review in both sentence and aspect level. A featured specific comparison has been made with the results that are retrieved using the QD technique and the product with the desired feature has been recommended to the user. The QD miner technique is able to classify user opinions with good precision and accuracy.