ABSTRACT

At present, researchers have conducted research and development on retail management systems from different perspectives. It was found that most chain franchise store management systems lack sufficient data analysis, store evaluation, and sales forecasting. Therefore, data analysis and sales forecasting of chain franchise stores are of great significance. This study aims to predict the sales revenue of a chain store based on its basic information. The data comes from a sports shoe chain store limited company. This study determined that significant predictive factors can predict the sales revenue of a chain store based on factors such as the type of chain store, the location of the chain store, the region where the chain store is located, product discounts, holiday promotions, and order quantities.This study adopts a cross industry data mining standard process for modeling, first cleaning the data, and then conducting feature analysis on the data to determine the important factors that affect sales revenue. Subsequently, the data was standardized and a prediction model was constructed using the random forest algorithm. The accuracy of the model reached 97%, meeting the expected target. Based on the constructed model, this study developed an application system for chain store data analysis and predictive management. This system can collect new data. At the same time, the system deployed data analysis and prediction models to conduct detailed analysis of the collected historical sales data, which can predict the total sales of a certain chain store in the next time window (month). Data analysts or sales management decision-makers can use this system to conduct comprehensive evaluations and predictive analysis of chain stores, thereby achieving accurate management decisions. According to IT experts’ evaluation, the average score of the system is 4.46, which belongs to the “very high level” category. The developed system has been confirmed to comply with the ISO 25010 software quality standard.

Keywords: Chain Stores, Random Forest Algorithms, Sales Forecasting, System Evaluation, Modeling
https://doi.org/ 10.57180/lfao4589