Experimental data for test the WPS tier and quality report: results of the quality control service

This dataset contains the results of the WPS tier and quality report presented in the thesis "Automatic evaluation of geospatial data quality using web services". The thesis' main goal was to develop a web service able to evaluate the quality of geospatial datasets using the standardized interface of Web Processing Service (WPS) in a fully automatic way. WPS is an open specification from the Open Geospatial Consortium (OGC) that defines an interface aiming to publish, use, discovery and binding geospatial processes. In the proposed architecture of the quality control service, the WPS tier plays the role of interface between clients and the evaluation procedures. This dataset contains the files used in the experiment to test the implementation of the WPS tier, which includes the quality reports.

In order to test the WPS tier, we ran three essays, one for each service developed: (w1) topological consistency service, (w2) completeness service, and (w3) positional accuracy (planimetry) service. These services implement quality evaluation procedures described in the Brazilian standard of geospatial data quality, named CQDG (http://www.geoportal.eb.mil.br/). We tested these services using data from the matching testbed (https://www.nature.com/articles/sdata2017180). These essays are summarized in Fig_Overview_of_requests.png.

The first essay, named W1, checked the topological consistency service. There are four requests: w1-01 to w1-04, with their correspondent responses.

The W2 essay evaluated the implementation of the completeness service. This essay has four data inputs: (1) Test: the test dataset (matching testbed S7P); (2) Reference: the reference dataset (matching testbed B7P); (3) ScaleDenominator: the scale of test dataset, used in sampling (1:25,000); and (4) SamplingStrategy: the sampling strategy to be adopted when calculating the commission and omission rates. There are six requests: w2-01 to w2-06, with their correspondent responses.

The last essay for the WPS tier, named W3, evaluated the implementation of the positional accuracy (planimetry) service, which accomplishes the automatic quality procedures for positional accuracy of planimetry in the Brazilian standard. The output values are: (1) CQDG:301: the value of quality category; (2) ER90: 90% percentile of planimetric errors; and (3) RMS: root mean square (RMS) of planimetric errors. They can be combined into a quality report in ISO 19157-2 XML format. There are 33 requests: w3-01 to w3-03, w3-040 to w3-049, w3-140 to w3-149, w3-240 to w3-249, with their correspondent responses. The first three requests used the dataset S1A (see matching testbed), and the last 30 requests used 10 datasets with random disturbance with a standard displacement of 5 m (see matching testbed).

本数据集收录自论文《基于Web服务的地理空间数据质量自动评估》中的WPS层级与质量评估报告结果。该论文的核心目标是开发一款Web服务，可通过Web处理服务（Web Processing Service, WPS）的标准化接口，以全自动化方式完成地理空间数据集的质量评估。WPS是开放地理空间信息联盟（Open Geospatial Consortium, OGC）制定的开放规范，定义了用于发布、调用、发现及绑定地理空间处理流程的接口标准。在本次提出的质量控制服务架构中，WPS层级承担了客户端与质量评估流程之间的接口角色。本数据集包含了用于测试WPS层级实现的实验相关文件，其中即包含各类质量评估报告。
为测试WPS层级，我们针对开发的三项服务分别开展了实验：(w1)拓扑一致性服务、(w2)数据完整性服务，以及(w3)平面位置精度（planimetry）服务。这些服务均实现了巴西地理空间数据质量标准CQDG（http://www.geoportal.eb.mil.br/）中规定的质量评估流程。我们采用匹配测试床（https://www.nature.com/articles/sdata2017180）提供的数据对上述服务进行了测试，相关实验概况汇总于Fig_Overview_of_requests.png。
第一项实验W1针对拓扑一致性服务展开，共包含w1-01至w1-04共4组请求及对应的响应结果。
第二项实验W2用于评估数据完整性服务的实现。该实验包含四类输入数据：(1) 测试数据集：匹配测试床S7P；(2) 参考数据集：匹配测试床B7P；(3) ScaleDenominator：测试数据集的比例尺，用于采样计算，取值为1:25,000；(4) SamplingStrategy：计算误报率与漏报率时采用的采样策略。本次实验共包含w2-01至w2-06共6组请求及对应的响应结果。
针对WPS层级的最后一项实验W3，用于评估平面位置精度（planimetry）服务的实现，该服务可按照巴西国家标准自动执行平面位置精度的质量评估流程。其输出结果包含三类：(1) CQDG:301：质量类别取值；(2) ER90：平面误差的90%分位数；(3) RMS：平面误差的均方根（root mean square, RMS）。上述结果可整合为符合ISO 19157-2 XML格式的质量报告。本次实验共包含33组请求：w3-01至w3-03、w3-040至w3-049、w3-140至w3-149、w3-240至w3-249，以及对应的响应结果。其中前3组请求采用S1A数据集（详见匹配测试床说明），剩余30组请求则采用10组带有随机偏移的数据集，偏移标准差为5米（详见匹配测试床说明）。