Latest Updates

Reviewers are welcome to join IRJTAS JOURNAL.
Authors are invited to publish their manuscript.
First edition is in process for publication.



Sentiment Analysis of Amazon Mobile Reviews Based on FeatureExtraction Approach over Cloud Environment

Kapil Jain Prof. Amit Ganguli Prof. Ajit Kumar Shrivastava

MTech Scholarkpljain21@gmail.com M. Tech. Co-ordinatoramitganguli@sistec.ac.inHead of Dept, CSESISTecR.hodcs@sistec.ac.in

ABSTRACT

Online shopping has been growing for 20 years and many e-commerce websites such as Amazon, have been created to meet the increasing demand. Consequently, a specific product can be bought on several websites and the prices may vary. As customers usually want the best quality for the lowest price but cant directly check it, reviews from other customers seem to be the most reliable way to decide whether to buy the product or not.Therefore, sentiment analysis has proven essential to understand a products popularity among the buyers all over the world. Sentiment analysis is a classification process whereby machine learning techniques are appliedon text-driven datasets in order to analyze its sentiment. But the data of ecommerce are growing rapidly and need high end processor to process these data. Distributed computing resembles a panacea to defeat the obstacles. Itvows to expand the speed with which the applications are conveyed, expanded imagination, development, brings down cost at the same time expanding business sharpness. It calls for fewer ventures and a collect of advantages. The end-clients just compensation for the measure of assetsthey utilize and can without much of a stretch scale up as their necessities develop. Specialist co-ops, then again, can use virtualization innovation to expand equipment use and work on administration. In this, we will take an virtual machine on cloud preferred over others because cloudservice is provided by third party providers (Google cloud platform (GCP)) so for security reason private cloud give better security than others because the connection between user and virtual machine is secured by ssh. And on private cloud we will easily scale the storage and processing powerat any time whenever application required. and that we can build logistic regression model supported different feature extraction techniques like BOW (Bag of Words), TF-IDF and N-Gram , From experimental result we willsay that model repose on N-Gram features provides better accuracy as compared to others.

Keywords : cloud computing, private cloud, machinelearning application, security, ssh, feature extraction, Bag ofWords, TFID, N-Gram.