Marketing Intelligence Through Consumer Review Analysis A Text Mining Approach

Abstract

quotBackground newlineIn the era of internet, several consumers are quite open to share their views about newlinegoods and services in various review portals. These reviews can be useful in newlineunderstanding how the product is doing in the market. Particularly, how competitive newlineproducts are being viewed in the market could be important piece of information for any newlinebusiness. newlineAim newlineThe current study is focused at extracting business relevant information from consumer newlinereviews which can be helpful in devising future marketing activities. In this regard, this newlinestudy is purely exploratory in nature. Text mining tools are used throughout the study to newlineunderstand what people are talking about particular products existing in the market. newlineMaterials and Methods newlineConsumer reviews were collected using web crawling techniques. The reviews were newlinecollected from flipkart.com as this e-retailer is very popular in India for selling various newlinegoods online. Moreover, it gives the consumers the opportunity to put forward their newlinereviews on goods or services which act as an easy reference for other buyers to decide newlinewhich product to buy. No sampling technique was used here because, for this study, all newlinethe available reviews were collected for the study. newlineSubhasis Dasgupta vi newlineResult and Discussion newlineThrough this study it was found out that the reviews could be used to create perceptual newlinemaps for studying market structure. The analysis of reviews also showed that battery newlineplays a critical role in deciding the overall rating of a mobile phone. The study also shed newlinesome light on how people compare one product with another, for example, Galaxy Note newlinewas mostly compared to Iphone 5S but Xperia Z was more compared to Galaxy Note. newlineMoreover, it was also found out that people talk about different topics while writing their newlinereviews on different products. The topics were different, mostly uncorrelated and newlineimportance of different features varied among those topics. newlineConclusion newlineIt can be concluded that using available text mining concepts, various interesting newlineinsights can be found out from p