|START Conference Manager|
IGN France has transformed its source data into the INSPIRE data models, according a wide-spread methodology: matching tables, transformation to pseudo-INSPIRE data set and on-the-fly transformation into GML by the WFS server (degree). The validation of data has two main objectives: to ensure conformity to INSPIRE (validation against target schema) and to ensure that the transformation has not degraded the data (validation against source data). The validation against schema is done using the INSPIRE validator. Until now, this tool has been used on-line. It is user-friendly, provides understandable error reports but it didn’t provide stable results (use of beta versions). The validation against source data includes the following steps: extraction of a sample test data from the WFS and then comparison between source data and target INSPIRE data. In order to carry out this comparison, a document lists the tests to be performed; this document is of course based on the transformation rules defined in the matching table. Then, the tests are designed as FME workbenches and run on sample data. An error report is generated and generally leads to some corrections in the transformation process. The validation against source data has proved to be quite useful as for each theme, some errors have been discovered, some minor (naming conventions), some major (missing features or errors in transformation rules). IGN has developed an automatic tool to extract sample data from WFS, providing us some concrete experience about use of our own INSPIRE services. The extraction of features with direct geometry is easy and works well on reasonable sample but the recuperation of the associated features (such as transport properties or address components) revealed to be very fastidious and effective only on very small number of features. Though done with great care, the matching tables have had to be brought some (minor) corrections during the initial transformation process or during the validation process. This has highlighted the need to maintain the matching tables.
Topic Area: [3.5] Best practices Abstract Type: Oral Presentation
Academic: No Data Provider: Yes Data User: No INSPIRE Implementer (IT): Yes INSPIRE newbies: No Policy Officers: No Public Administration (MS/Regional/Local): No Thematic specialists: No
START Conference Manager (V2.61.0 - Rev. 5269)