Sentiment Analysis Using Hadoop Framework and Machine Learning Approach
Ritu Patidar, Sachin Patel · 2023
Opinion mining is the study of user opinions as revealed by their text messages. It comprises categorizing user attitudes into several polarities, such as positive, negative,or neutral. For the analysis, a whole other framework is required, one that can process the enormous volume of data quickly and accurately while maintaining a high levelof unpredictability. More than half of the data produced by e-commerce platforms like Amazon, Flipkart, and others come in the form of text, amounting to 20ZB. These text messages can be carefully analyzed and studied to gaina clear understanding of the thoughts and opinions regarding every part of the business. Over the past ten years, analyzing this enormous amount of data and forecasting user behavior have been the major challenges. To analyze the attitudes, we combine the Hadoop infrastructure with the machine learning technique in this study. With the proposed architecture we will be evaluated certain metricsparameters. It is observed that the random forest achieves better accuracy 96 percent as compared to other machine learning algorithms, hence it works efficiently in Big data environments and achieves maximum accurate results.