Visualization of Spatio-Temporal Events in Geo-Tagged Social Media
Visualization of Spatio-Temporal Events in Geo-Tagged Social Media
复制标题
DOI：
10.1007/978-3-319-55998-8_9
复制
发表时间：
2017-05
期刊：
影响因子：
--
通讯作者：
Yuanyuan Wang;Muhammad Syafiq Mohd Pozi;Goki Yasui;Yukiko Kawai;K. Sumiya;Toyokazu Akiyama
中科院分区：
文献类型：
--
作者：
Yuanyuan Wang;Muhammad Syafiq Mohd Pozi;Goki Yasui;Yukiko Kawai;K. Sumiya;Toyokazu Akiyama
本文提出了一种时空映射系统，用于将地理标记的社交媒体摘要可视化为云中的标签，并通过检测时空事件与网页相关联。通过它，用户可以随时随地在浏览任何网页时掌握事件。为了从推特等社交媒体中检测时空事件，系统利用机器学习算法通过空间和时间对推特进行分类，提取预期事件（如拥挤的餐馆），并将当前情况与正常规律进行比较，提取意外事件或季节性事件（如时间销售）。因此，系统呈现tweet的社交标签云，帮助用户在浏览网页时快速了解事件的时空概况，并呈现tweet列表，帮助用户获取事件的更多细节。此外，用户可以自由地指定一个时间段或一个标签来查看其相关的tweets。最后，我们讨论了我们提出的社会标签云生成方法的有效性，使用密集的地理标记推文在城市地区的多功能建筑。
This paper presents a spatio-temporal mapping system for visualizing a summary of geo-tagged social media as tags in a cloud, and it is associated with a web page by detecting spatio-temporal events. Through it, users can grasp events at anytime from anywhere while they browse any web pages. In order to detect spatio-temporal events from social media such as tweets, the system extracts expected events (eg, crowded restaurants) by using machine learning algorithms to classify tweets through space and time, and it also extracts unexpected or seasonal events (eg, time sales) by comparing the current situation to those normal regularities. Thus, the system presents a social tag cloud of tweets to help users gain a quick overview of spatio-temporal events while they browse a web page, and it also presents a tweet list to help users obtain more details about events. Furthermore, users can freely specify a time period or a tag to view its related tweets. Finally, we discuss our proposed social tag cloud generation method’s effectiveness using dense geo-tagged tweets at multi-functional buildings in urban areas.