Abstract:
"At present the concept of e-WOM is heavily used by the businesses, e-retailing managers, and marketers to create awareness of their products and services for a greater audience. Having more than 186 million daily active users, Twitter has become one of the major social media networks.
Accordingly, the purpose of this research is to find out the tweet related factors that would positively impact on e-WOM. This study will measure the virality of a tweet by the number of retweets received for a particular tweet as retweeting indicates that the tweet post has been spread among many users.
The relationship between tweet related factors namely, hashtag, mention, media, emoji, length of the content, symbols and link have been considered to analyze the influence of the content using 76594 tweets from three of the top cosmetic brands in USA that heavily use social media as a platform to conduct their marketing activities.
In this study several attributes extracted from Twitter data have been analyzed using SVM, Logistic regression, Decision tree, Naïve bayes, Random Forest and XG Boost machine learning algorithms to predict that factors that are mostly impacted on retweets. Further, the analysis is conducted using R Programming language.
Findings of this study would help e-retailing marketers and managers revamp their marketing techniques to induce retweeting which will in return positively impact on the process of e-WOM
"
