By recognizing the decision based on the reviews of the customer behavior, the decision-making of the customer who purchases the product in the e-commerce platform has been enabled. The system could not make decisions due to a review difficulty from a specific site or application. This research regression tree algorithm employs the automated decision-making system technique for decision making. The KNN is also used to collect data from reviews. The automated decision-making system is utilized in the decision-making process for the system's classification to produce customer behavior. In this case, the KNN is employed to classify the data, followed by controlling the automated decision-making system based on the reviews on the various platforms. The online platform has increased 90% of the purchasing history in 2019–2020 compared to 2017–2018. When comparing 2019–2020 to 2020–2022, the internet platform accounted for 100% of all purchases. Using this variation, the online purchase has managed the factor for an online platform for the automated decision-making system over the last two years. The suggested system has a 0.94 accuracy, 0.96 precision, 0.95 recall, and 0.97 F-measure. These numerical measurements demonstrate how well the system can evaluate and forecast client behavior. In addition, the suggested approach effectively decreased return rates across a range of product categories, including electronics (16.8% reduction), clothes (11.2% reduction), home and a kitchen (12.7% reduction), beauty and personal care (10.4% reduction), and sports and fitness (17.3% reduction) and up to 96% customer retention in month twelve. These outcomes demonstrate significant operational and strategic worth.
Panga et al. (Thu,) studied this question.