aurman/GoogleTrendArchive

该数据集包含超过1020万个来自Google“Trending Now”功能的实时搜索趋势实例，数据持续收集于2024年11月28日至2026年5月17日期间，覆盖所有可用的地理位置（200多个国家和地区）。与Google Trends等聚合回顾工具不同，Trending Now捕获了正在经历实时激增的搜索查询，为研究集体注意力动态提供了前所未有的时间粒度。每个实例代表特定搜索词或查询集群根据Google算法成为“趋势”的时刻，包含搜索量桶、精确时间戳、趋势持续时间、地理位置和相关查询变体。数据集由苏黎世大学社会计算组和斯坦福人工智能实验室的Aleksandra Urman、Anikó Hannák和Joachim Baumann策划，语言为多语言，许可证为CC-BY-4.0。This dataset contains over 10.2 million real-time search trend instances sourced from Google's "Trending Now" feature. The data was continuously collected between November 28, 2024 and May 17, 2026, covering all available geographic regions (over 200 countries and territories). Unlike aggregated retrospective tools such as Google Trends, "Trending Now" captures search queries experiencing real-time surges, providing unprecedented temporal granularity for research on collective attention dynamics. Each instance represents the moment when a specific search term or query cluster is classified as a "trend" by Google's algorithm, and includes search volume bins, precise timestamps, trend durations, geographic locations, and related query variants. The dataset was curated by Aleksandra Urman, Anikó Hannák and Joachim Baumann from the Social Computing Group at the University of Zurich and the Stanford Artificial Intelligence Laboratory. It is multilingual and licensed under CC-BY-4.0.