Detection of Reviews and Comments using Machine Learning

Abstract

The modern internet and different sources are the best resources for knowledge, inspiration, and product or service reviews. Millions of comments about a product, a person, or a location are made every day in a variety of settings or sources. It is exceedingly challenging to manage and comprehend such comments due to their enormous volume and size. The user's viewpoint is always expressed in textual form. To identify whether the opinion's tone is favourable, unfavourable, or neutral, artificial intelligence and text analytics are applied. The feedback is gathered from an internet source, then analysed and saved in a databases. Those comments are identified and classified whether it was user keyword related post using naive bayes algorithm. The phrase “Naive Bayes Classification” comes from the premise that each word is statistically independent of the others. The user keywords can be predicted whether it is a best suggestion using polarity. To develop an interactive autom system that can anticipate the emotion of comments. This system talks about the challenges that can come u during sentiment analysis. Comments are taken into consideration because they offer rich data sources for sentiment analysis and opinion mining. The main objective of this system is to analyse gathered comments for emotional performance in real-time and offer time-based statistics to the user via an automated teller machine.