QUALITY OVER QUANTITY IN GLOBAL E-COMMERCE: DECODING PERFORMANCE THROUGH EXPLAINABLE MACHINE LEARNING AND STRATEGIC SIMULATIONS

This dataset contains the performance and user interaction metrics for the 50 most visited e-commerce platforms globally as of June 2025. It was compiled to support the research article titled "Quality Over Quantity in Global E-Commerce: Decoding Performance Through Explainable Machine Learning and Strategic Simulations".
The data includes nine key performance indicators: Average Visit Duration, Page Visits, Bounce Rate, Performance Coefficient, Speed Index, Unique Visitors, Monthly Visits, Traffic Share, and Month-over-Month (MoM) Traffic Change.
Data Collection Methodology: The raw data was collected from reputable digital analytics tools, including SimilarWeb, GTmetrix, and Google PageSpeed Insights. The dataset allows for the replication of the hybrid methodology involving CRITIC-based weighting, multiple MCDM methods (TOPSIS, VIKOR, etc.), and machine learning predictions using SHAP analysis.

本数据集收录了截至2025年6月全球访问量排名前50的电子商务平台的性能与用户交互指标，旨在支撑题为《全球电子商务领域以质胜量：通过可解释机器学习与战略模拟解析平台性能》的研究论文。
本数据集包含九项核心绩效指标：平均访问时长、页面访问量、跳出率、性能系数、速度指数、独立访客数、月度访问量、流量占比及月度环比（Month-over-Month, MoM）流量变化。
数据采集方法：原始数据采集自SimilarWeb、GTmetrix、Google PageSpeed Insights等权威数字分析工具。本数据集支持复现包含基于CRITIC权重法、多种多准则决策（Multi-Criteria Decision Making, MCDM）方法（TOPSIS、VIKOR等）以及采用SHAP分析的机器学习预测在内的混合研究方法。