media-metadata-steam-games
收藏资源简介:
Steam Games 数据集是一个包含约82,000款PC游戏信息的集合,数据来源于Steam公共商店API,涵盖了Steam完整的游戏目录(不包括DLC、演示版本和工具软件)。该数据集旨在提供全面的游戏元数据,包括定价、玩家评论情感分析、所有权估计以及峰值并发用户数等关键指标,适用于游戏市场分析、流行度排名、推荐系统研究等任务。数据集包含12个结构化字段:Steam应用ID(steam_appid)、游戏名称(name)、开发者(developer)、发行商(publisher)、正面评论数(positive_reviews)、负面评论数(negative_reviews)、所有权区间估计(owners,例如“1,000,000 .. 2,000,000”)、所有者的平均游戏时长(average_playtime_forever,单位:分钟)、所有者的中位游戏时长(median_playtime_forever,单位:分钟)、当前美元价格(price_usd,0.0表示免费)、当前折扣百分比(discount_pct)、峰值并发用户数(ccu,抓取时的数据)以及Steam游戏类型标签(genres)。其中,峰值并发用户数和评论数量可用于评估游戏实时流行度;所有权估计字段采用了类似Gamalytic的基于SteamSpy的区间估计方法。数据集以表格形式组织,便于直接加载和分析。
The Steam Games dataset is a collection of approximately 82,000 PC game information, sourced from the Steam public store API, covering the complete Steam game catalog (excluding DLC, demos, and tool software). This dataset aims to provide comprehensive game metadata, including key metrics such as pricing, player review sentiment analysis, ownership estimates, and peak concurrent user counts, suitable for tasks like game market analysis, popularity ranking, and recommendation system research. The dataset contains 12 structured fields: Steam application ID (steam_appid), game name (name), developer (developer), publisher (publisher), positive review count (positive_reviews), negative review count (negative_reviews), ownership range estimate (owners, e.g., 1,000,000 .. 2,000,000), average playtime per owner (average_playtime_forever, in minutes), median playtime per owner (median_playtime_forever, in minutes), current USD price (price_usd, 0.0 indicates free), current discount percentage (discount_pct), peak concurrent users (ccu, data at the time of scraping), and Steam genre tags (genres). Among these, peak concurrent user counts and review numbers can be used to assess real-time game popularity; the ownership estimate field uses a range estimation method based on SteamSpy, similar to Gamalytic. The dataset is organized in a tabular format for easy loading and analysis.




