Pre-Pandemic E-Commerce Patterns

Ashley Pacheco

Co-Presenters: Joseph Margaritondo, Habiba Morsy

College: The Dorothy and George Hennings College of Science, Mathematics and Technology

Major: Computational Science & Engineering - STEM 5 Year B.S./M.S.

Faculty Research Mentor: Ching-yu Huang

Abstract:

The rapid growth of e-commerce has transformed global consumer behavior due to the Covid-19 pandemic. This study analyzes e-commerce patterns between 2011 and 2015, focusing on pre-pandemic trends across different countries. Using the Global SuperStore Dataset from Kaggle, which has a raw data size of 7,848 KB containing 51,290 records and 24 attributes, we examine raw data factors such as products, shipping costs, profit, and discount effectiveness to determine demand patterns and profitability across regions. The dataset has been cleaned and processed to enhance analysis, with additional engineered features such as Delivery Time and Aggregated Sales & Profit to improve insights. The final dataset is stored either as a CSV file for further analysis or in a MySQL database for querying. This structured data allows for effective application of data mining techniques, helping to identify trends in e-commerce sales, customer preferences, and regional profitability. After preprocessing is complete, the dataset is analyzed using both data mining and statistical techniques. Clustering is used to group customers from different countries and identify spending habits, and regression is used to make predictions about future profits and types of customer purchases. Furthermore, measures of central tendency such as mean and median are calculated to provide further insight into the distribution of data points. With this analysis, this study expects to find that discounts do indeed generate more sales and overall profit, despite the lowered profit per individual sale. It is also expected that favored categories of purchases exist and are discernable from each country described in the dataset, though no prediction is made on what those particular categories are. This research provides insights into consumer preferences, regional sales trends, and logistical efficiencies, offering valuable information for businesses seeking to optimize their e-commerce strategies.