FISETIO: A FIne-grained, Structured and Enriched Tourism Dataset for Indoor and Outdoor attractions

This data in brief paper introduces our publicly available datasets in the area of tourism demand prediction for future experiments and comparisons. Most previous works in the area of tourism demand forecasting are based on coarse- grained analysis (level of countries or regions) and there are very few works and datasets available for fine-grained tourism analysis as well (level of attractions and points of interest). In this article, we present our fine-grained datasets for two types of attractions – (I) indoor attractions (27 Museums and Galleries in U.K.) and (II) outdoor attractions (76 U.S. National Parks) enriched with official number of visits, social media reviews and environmental data for each of them. In addition, the complete analysis of prediction results, methodology and exploited models, features’ performance analysis, anomalies, etc, are available in our original paper本简要论文介绍了我们公开的旅游需求预测领域数据集，以供后续实验与对比研究使用。此前该领域的旅游需求预测研究大多基于粗粒度分析（国家或区域层面），针对景点与兴趣点层面的细粒度旅游分析相关研究与可用数据集却十分匮乏。本文中，我们公布了两类景点的细粒度数据集：(I) 室内景点（英国境内27家博物馆与美术馆）；(II) 室外景点（美国境内76座国家公园），数据集补充了各景点的官方到访人次、社交媒体评论与环境数据。此外，关于预测结果、研究方法与所采用模型的完整分析、特征性能评估、异常情况分析等内容均可在我们的原始论文中查阅。