Optimizing Classification Methods for Online Buyers' Purchase Intentions in Bangladesh

Abstract

The classification of online buyers' purchasing intentions is of paramount importance, especially in the context of the period of the COVID-19/post-COVID-19 pandemic, as it carries significant implications for the business industry. However, effectively managing the diverse ever-changing intentions of individual Internet customers remains a challenging task. This study aims to improve the classification techniques used to classify different sorts of online buyers' purchasing intents in Bangladesh. A comprehensive analysis of different classification algorithms reveals that the Random Forest algorithm outperformed other methods, achieving exceptional accuracy rates of 99.9% in training and 89.7% in testing. Conversely, the Gaussian Naive Bayes algorithm demonstrated comparatively lower accuracy, with training testing accuracies of 80% and 79%, respectively. This study contributes not only to a better understanding of online buyers' purchase intentions in Bangladesh but also provides valuable insights into the business industry. Moreover, our work highlights the potential for future investigations in recognizing Bangla numerals throug gestures to enhance the accuracy of categorizing online buyers' intended purchases. This research serves as a stepping stone for further advancements in classifying and understanding online buyers' purchase intentions, ultimately fostering more accurate decision-making in the realm of E-commerce in Bangladesh.