Retail Dataset

Dataset Description

This dataset contains 436,689 records of commercial transactions from a retail store/marketplace, recorded between December 2014 and December 2015. Each row represents the purchase of a single product within a specific invoice.

Main Columns:

1. InvoiceID: Unique identifier for the invoice (categorical).
2. CustomerID: Unique identifier for the customer (categorical).
3. date: Date of the transaction (temporal), useful for time series and seasonality analysis.
4. item: Product description (text/categorical).
5. quantity: Quantity purchased for the individual product (numeric, discrete).
6. price: Unit price of the product (numeric, continuous).
7. type: Sales channel (e.g., online, supermarket), categorical.
8. category: Product category, useful for grouping and pattern mining (categorical).
9. total_quantity: Total quantity of products purchased in the same invoice (numeric, discrete).
10. customer_type: Type of customer (e.g., private, wholesaler), categorical.
11. product_id: Unique identifier for the product (categorical).

Potential Data Mining Applications:

• Market Basket Analysis / Association Rules: Discover frequent combinations of products purchased together.
• Customer Segmentation: Cluster customers based on purchase quantity, customer type, and product categories.
• Forecasting: Predict future sales for products or categories.
• Anomaly Detection: Detect unusual transactions in terms of quantity or price.
• Sales Analysis: Identify trends, top-selling products, and seasonal patterns.

General Characteristics:

• Large dataset with a mix of online and in-store sales.
• Combination of numeric and categorical variables, suitable for classification, clustering, and association rule mining.
• Includes a temporal variable for sequential and predictive analysis.