社交电商数据集

本数据集包含100,000条社交电商用户购买行为记录，涵盖31个特征变量和1个二分类目标变量（是否购买）。小红书、抖音等社交电商平台的真实场景，包含用户特征（年龄、性别、等级等10个）、内容特征（价格、折扣、类目等7个）、社交特征（点赞、评论、分享等6个）、行为序列特征（加购、用券、浏览等5个）以及4个衍生特征（互动率、购买意向等）。数据集正负样本比例约为1:4，用户以年轻女性为主（平均年龄27岁，女性占比63.8%），价格和互动数据呈右偏分布，符合社交电商的典型特征，适用于购买转化预测、推荐系统优化、用户行为分析等机器学习任务。

This dataset contains 100,000 user purchase behavior records from social e-commerce platforms, including 31 feature variables and 1 binary classification target variable (whether a purchase was made). The data is collected from real-world scenarios of mainstream social e-commerce platforms such as Xiaohongshu (Little Red Book) and Douyin (TikTok). The 31 feature variables are categorized into five groups: 10 user-related features (age, gender, user level, etc.), 7 content-related features (price, discount, product category, etc.), 6 social interaction features (likes, comments, shares, etc.), 5 behavioral sequence features (shopping cart addition, coupon redemption, browsing, etc.), and 4 derived features (interaction rate, purchase intention, etc.). The positive-to-negative sample ratio of the dataset is approximately 1:4. The user group is dominated by young females, with an average age of 27 and females accounting for 63.8% of the total. Price and interaction data follow a right-skewed distribution, which aligns with the typical characteristics of social e-commerce. This dataset is suitable for machine learning tasks including purchase conversion prediction, recommendation system optimization, and user behavior analysis.

搜集汇总

背景与挑战

背景概述

该数据集是一个包含10万条社交电商用户购买行为记录的公共数据集，涵盖用户、内容、社交、行为序列和衍生特征共31个变量，目标变量为是否购买的二分类标签。数据集以年轻女性用户为主，正负样本比例约为1:4，价格和互动数据呈右偏分布，适用于购买转化预测、推荐系统优化等机器学习任务，真实反映了小红书、抖音等社交电商平台的典型场景。