Integrated Multi-Source, Multi-Country Retail Sales Dataset with Engineered Seasonal Features

This dataset is a processed and integrated retail sales dataset constructed from multiple public and private data sources across different regions. It includes data from the United Kingdom, Malaysia, and selected European countries (France, Spain, Portugal, Germany), combining both real-world and synthetic retail datasets.

The dataset contains approximately 70,000 records and provides transaction-level and aggregated sales information, including date, product, category, price, quantity, and revenue. Data from heterogeneous sources were cleaned, standardized, and merged into a unified schema to ensure structural consistency across different datasets.

Additional temporal and seasonal features were engineered to support time-series analysis, including weekday and weekend indicators, month identifiers, day-of-week values, country-specific holiday flags, and seasonal labels. These features capture both calendar-based effects such as holidays & seasonal cycles and region-specific consumption patterns.

The dataset is designed to represent diverse retail environments and heterogeneous data conditions, including variations in product categories, regional behavior, and data granularity. It can be reused for tasks such as time-series modeling, feature engineering, benchmarking, and cross-regional analysis under varying seasonal conditions.

本数据集为经处理整合的零售销售数据集，依托跨多区域的多个公开与私有数据源构建而成。其数据来源覆盖英国、马来西亚以及部分欧洲国家（法国、西班牙、葡萄牙、德国），融合了真实零售数据集与合成零售数据集。

本数据集包含约7万条记录，提供交易级与聚合级销售信息，涵盖日期、商品、品类、售价、销量与营收等字段。针对多源异构数据，我们完成了清洗、标准化操作，并将其整合至统一数据模式中，以确保不同来源数据集间的结构一致性。

此外，我们还工程化构建了额外的时序与季节特征以支撑时序分析任务，具体包括工作日与周末标识、月份标识、星期几取值、国别专属假日标记，以及季节标签。这些特征可捕捉基于日历的假日与季节周期效应，以及区域特有的消费模式。

本数据集旨在表征多样化的零售场景与异构数据条件，涵盖商品品类、区域消费行为与数据粒度等多维度差异。其可复用于时序建模、特征工程、基准测试，以及不同季节条件下的跨区域分析等任务。