Abstract
Online customer reviews are important sources of information influencing consumers’ attitudes towards products and brands. Businesses also use them to gain deeper insights into consumers’ perceptions, attitudes, and behaviors. This study uses machine learning (ML) and participant-based attributes of reviewers to classify them into distinct segments. The segments are then labelled and used to build a predictive model of customer behavior, which can help companies quickly profile customers and develop appropriate marketing strategies. The results show that our machine learning approach coupled with participation-based attributes generated unique clusters that are consistent with prior classification of online audiences. The personas-based clusters can help marketers make better use of reviewers in marketing campaigns by engaging them differently based on their interests and status in the online community. This study opens the door for further research using ML with larger and different review sites coupled with additional psychological, social, and economic variables.
Recommended Citation
Darbanibasmanj, Amir Abbas; Persaud, Ajax; and Ruhi, Umar, "Application of Machine Learning to Mining Customer Reviews" (2019). AMCIS 2019 Proceedings. 21.
https://aisel.aisnet.org/amcis2019/human_computer_interact/human_computer_interact/21
Application of Machine Learning to Mining Customer Reviews
Online customer reviews are important sources of information influencing consumers’ attitudes towards products and brands. Businesses also use them to gain deeper insights into consumers’ perceptions, attitudes, and behaviors. This study uses machine learning (ML) and participant-based attributes of reviewers to classify them into distinct segments. The segments are then labelled and used to build a predictive model of customer behavior, which can help companies quickly profile customers and develop appropriate marketing strategies. The results show that our machine learning approach coupled with participation-based attributes generated unique clusters that are consistent with prior classification of online audiences. The personas-based clusters can help marketers make better use of reviewers in marketing campaigns by engaging them differently based on their interests and status in the online community. This study opens the door for further research using ML with larger and different review sites coupled with additional psychological, social, and economic variables.