dataset for multiple LBSs

Yelp is a typical one-platform dataset, with the last update at February 16 2021 for Yelp Dataset Challenge. Here, the review service is regarded as the check-in service because the review data record online interaction between users, whose fields include the useful count, the funny count and the cool count. We filter for users whose city is Austin, delete the records with missing values, and extract a subset of 45,524 users with 12,855 locations.
The foursSwarm dataset is obtained through the alternative crawl, whose user trajectories compose tips from Foursquare and check-ins from Swarm. In this dataset, locations are generated from October 9 2008 to November 5 2017, and are filtered by geographical distance within the fixed location number, i.e., 10,000 in this paper. After data was preprocessed, we eventually extracted a subset of 6,982 users with 12,503 locations.

Yelp 数据集系一平台数据集之典范，其最新更新日期为2021年2月16日，针对Yelp Dataset Challenge。在该数据集中，评论服务被视为签到服务，因为评论数据记录了用户之间的在线互动，其字段包括有用次数、有趣次数和酷炫次数。我们对位于奥斯汀市的用户进行筛选，删除了缺失值记录，并提取了包含45,524位用户及其12,855个位置的子集。foursSwarm 数据集通过替代爬取获得，其中用户轨迹由Foursquare 的贴士和Swarm 的签到组成。在此数据集中，位置信息生成于2008年10月9日至2017年11月5日之间，并通过固定位置数量内的地理距离进行筛选，即本文中的10,000个位置。经过数据预处理后，我们最终提取了包含6,982位用户及其12,503个位置的子集。