Future sales prediction of OnePlus using Amazon review’s data

Abstract

In this research, we predicted sales of OnePlus cell phone models using Amazon reviews information extracted from Amazon.in. Starting with extraction, classifying and then cleaning of data, particularly it’s showed that how different web and text mining techniques can be applied to organise this unstructured text data in a numerical and computable format which further can be used in Algorithmic models in R. For prediction we have used multiple regression models as MLR, SVM and TS prediction, the resultant data of each regression model is then compared and depending on these results the best model is selected. Additionally, using Tableau we presented results in the form of line/bar graphs of each model and how the end results are different when compared based on sales, month, year and as a whole. The result is more targeted on the sales prediction of any product sold on amazon and reviewed by customers on Amazon.