ABSTRACT

Open government promises increased transparency by providing its citizens datasets about city processes. Open data portals have been emerging all over the world as mines of open geographic datasets. Thematic web maps are key to understanding these open geographic datasets. Current thematic web maps are created by programmers and/or cartographers, and thus are not designed to be easily reused with new geographic datasets. As a result, they pose several challenges to non-experts wanting to adapt them to new scenarios. This article introduces a semi-automatic approach for the creation of thematic web maps by and for users with no prior training in cartography. The approach relies on the mapping between Stevens’ data types and Bertin’s visual variables, to suggest (meaningful) thematic map visualizations for a given input geographic dataset. It was implemented as a web prototype in AngularJS and evaluated with 19 participants. Results from the user study suggest that despite facing a few challenges in accurately identifying Stevens’ data types, participants managed to successfully create web maps and correctly answer spatial questions. The prototype and insights gathered from the user study are relevant to making cartographic products more accessible to a broader population, and open geographic data more usable in the context of an open government.

摘要

开放政府承诺通过提供有关城市进程的公民数据集来提高透明度。开放数据门户网站已在世界范围内兴起，成为开放地理数据集的地雷。专题网络地图是了解这些开放的地理数据集的关键。当前的专题网络地图是由程序员和/或制图师创建的，因此不能设计为易于与新的地理数据集一起重复使用。结果，它们给想要使他们适应新场景的非专家提出了一些挑战。本文介绍了一种半自动方法，供没有经过制图培训的用户以及由用户创建专题网络地图。该方法依赖于Stevens数据类型与Bertin视觉变量之间的映射，以针对给定的输入地理数据集建议（有意义的）主题地图可视化。它在AngularJS中作为Web原型实现，并有19位参与者进行了评估。用户研究的结果表明，尽管在准确识别Stevens的数据类型方面面临一些挑战，但参与者还是成功地创建了网络地图并正确回答了空间问题。从用户研究中收集的原型和见解与使制图产品更广泛地可供人们使用以及在开放政府的背景下更开放地域数据有关。