INSPIRE Data Specification Extensions
  1. Introduction
  2. Results of the Survey
    1. Purpose
    2. Methodology
    3. Overview
    4. Expertise and Background
    5. Motivation and Scope
    6. Challenges Reported
    7. Best Practices
  3. Inventory of Model Extensions
  4. The INSPIRE Model-Driven Methodology
  5. The Extension Methodology
  6. The Pattern Catalogue
  7. An End-to-End Tutorial Project
  8. Conclusions and Outlook

Purpose

The first step in this project was to conduct a study, which had two goals:

  1. Create an inventory of existing models that relate to INSPIRE data specifications
  2. Understand what information this project would need to provide to its audience by asking about people’s expertise, lessons learnt and challenges encountered

The survey also had the side-effect of raising awareness, both of the possibility to extend INSPIRE models and of motivating experts to contribute to the effort.

Methodology

To reach these two objectives, we decided to go with a classical survey that would take 5 to 15 minutes to complete. The final survey contained 21 questions in four sections:

  1. Expertise
  2. Existing Models
  3. Planned Models
  4. Involvement and Contact Info

Several of the questions were conditional and needed to be answered only if a specific answer was given previously. We implemented the questionnaire using Google Forms. The link to the questionnaire was distributed via different channels to ensure a good coverage of the INSPIRE community. These channels included:

  • Wetransform mailing list and news section
  • LinkedIn INSPIRE Interest Group
  • LinkedIn personal and company updates
  • Sharing through Twitter
  • Direct communication with known experts and INSPIRE Maintenance and Implementation Group members

Overview of Responses

Between March 20th and May 31st 2016, 112 people responded to the survey. The overall response rate is estimated at 3%. Some channels, such as the direct mails to the INSPIRE Maintenance and Implementation Groups, naturally had a higher response rate (between 10 and 20%). The rate of incomplete responses was relatively low, at 3%.

The questionnaire is still online and open. We will update the numbers and figures on this page once per month if new entries are made, until the end of the project. The 112 respondents documented 34 existing data models related to the INSPIRE data specifications. 29 of documented an intent to create such a model in the near future.

Expertise and Background

Question 1.1: In what areas do you consider yourself an expert?
105 responses, multiple answers allowed

The majority of respondents to this survey are experts in geospatial data anaylsis, modeling and visualisation.

Question 1.2: How knowledgeable are you about standards for geospatial data and services?
109 responses

We have reached out to general practitioners but also to groups of experts on INSPIRE and ISO, and OGC standards, and this clearly shows in the responses.

Question 1.4: Have you worked with any INSPIRE Data Specification or a GML Application Schema?
95 responses

Question 1.5: Did you create at least one data model that references or extends INSPIRE data specifications, e.g. by including properties that are references to INSPIRE objects?
112 responses

Motivation, Scope and Usage

Question 2.1: Why did you create this data model?
36 responses

Question 2.3: Compared to the INSPIRE Data Specifications, what have you done to create your data model?
36 responses

Question 2.4: What methods did you use to relate your data model to other models?
35 responses

Question 2.6: How many people were involved in the creation of the data model?
36 responses

Question 2.7: Which schema language did you use to create the new data model?
36 responses

Question 2.9: Is the data model used?
36 responses

Question 2.11: How many versions of your data model have you released?
36 responses

Question 2.12: Will you continue to develop the data model?
34 responses

Question 2.13: What is the license of your data model?
33 responses

Challenges reported

Question 2.17: Which challenges did you encounter when designing the data model?
35 responses

Question 2.18: Which challenges did you encounter when implementing the data model?
35 responses

Best Practices

Question 2.19: Which best practices have you learned from creating the model? Please include links to any examples or documentation.
11 responses - 6 selected and anonymised for inclusion

  • [..] specifications were supposed to be build upon INSPIRE, some themes were extended due to existing products from [...], mostly due to requirements from [...]. For Planned Land Use, a mapping from the national models to INSPIRE was not so difficult, but it required an in depth study with a UML modeller and a Planned Land Use expert. As expected, the national model has more and stronger semantics. In general, the INSPIRE specification was fit for purpose.
  • Best practice is that it is possible and quite easy, yet powerful. In the next half year I'm focusing on creating extensions to CP. Next target is transforming models into RDF.
  • Learning INSPIRE schemas in practice revealed that extensions are possible and do not require extraordinary skills or resources
  • Involve stakeholders needs in data modelling and develop pilot projects
  • To reach a consensus by a heterogeneous group, flexibility of INSPIRE UMLs, interconnection between different INSPIRE themes at UML level, database implementation from INSPIRE data model, population of "INSPIRE" databases, INSPIRE GML creation & validation experiencies
  • UML Extending by mix-in. INSPIRE codelist register extension. RDF SKOS publication. Validation against codelist to be developed.