Synthetic Dataset and Analysis Code for an Integrated Naive Bayes Business Intelligence Approach to Maize Sales, Customer Satisfaction, and Marketing Performance

This repository provides all materials required to reproduce the analytical workflow presented in the study on a Naive Bayes-based Business Intelligence framework for maize sales, customer satisfaction assessment, and marketing performance evaluation. The contents include a fully synthetic dataset representing maize transaction activities, customer profiles, and structured responses to a multi-dimensional satisfaction instrument. The dataset was generated programmatically using a fixed seed to ensure full reproducibility of results and to support methodological validation without involving any real-world customer information. Included in this repository are: simulated transaction data covering sales, pricing, delivery, quality, and complaint attributes, survey data capturing customer satisfaction across multiple evaluation dimensions, scripts for synthetic data generation and preprocessing, implementation of the Naive Bayes classification model, evaluation outputs such as classification metrics, confusion matrix, and validation results, simulated marketing intervention outcomes including comparative performance analysis. The primary purpose of this repository is to demonstrate an end-to-end business intelligence pipeline, starting from data construction, measurement modeling, predictive analytics, to decision-oriented evaluation. It is intended strictly for academic and methodological illustration. All datasets are artificially generated and do not reflect actual customers, organizations, or operational records. Therefore, the outputs should not be interpreted as empirical findings from real field conditions. Any future application to real data must comply with ethical approval processes and data protection standards.