An Urban Data Profiler
WWW '15: 24th International World Wide Web Conference Florence Italy May, 2015(2015)
摘要
Large volumes of urban data are being made available through a variety of open portals. Besides promoting transparency, these data can bring benefits to government, science, citizens and industry. It is no longer a fantasy to ask "if you could know anything about a city, what do you want to know" and to ponder what could be done with that information. However, the great number and variety of datasets creates a new challenge: how to find relevant datasets. While existing portals provide search interfaces, these are often limited to keyword searches over the limited metadata associated each dataset, for example, attribute names and textual description. In this paper, we present a new tool, UrbanProfiler, that automatically extracts detailed information from datasets. This information includes attribute types, value distributions, and geographical information, which can be used to support complex search queries as well as visualizations that help users explore and obtain insight into the contents of a data collection. Besides describing the tool and its implementation, we present case studies that illustrate how the tool was used to explore a large open urban data repository.
更多查看译文
关键词
Metadata Extractionl,Automatic Type Detection,Dataset Analysis
AI 理解论文
溯源树
样例
生成溯源树,研究论文发展脉络