A Novel Approach for Predicting the Adoption of Smartwatches Using Machine Learning Algorithms

Abstract

The essential purpose of this research is to examine the behavioral intention to adopt smartwatches by individuals. A theoretical model was developed by integrating the Technology Acceptance Model (TAM) with the common smartwatch features, including mobility, availability, and perceived enjoyment. A quantitative approach using an online survey was employed to collect data from 491 students enrolled at seven higher educational institutes in Malaysia. This study employs a novel approach for testing the developed research model using six popular machine learning classification algorithms, including a meta classifier (LogitBoost), a logistic regression classifier (SimpleLogistic), a lazy classifier (KStar), a Bayesian classifier (NaiveBayes), a rule learner (OneR), and a decision tree classifier (LMT). The results indicated that SimpleLogistic and LMT performed better than the other classifiers in predicting the behavioral intention with an accuracy of 61%. This study has a novel contribution to the Information Systems (IS) literature since it was the first attempt to apply machine learning techniques in predicting the individuals’ adoption of smartwatches. The limitations and future research directions are also discussed.