The 52°North WPS implementation shares one important property with the specification – it is very flexible and can be used for many different use cases. A number of parties contribute, which results in a number of active development branches for code. We present a selection in blog posts in the series “A look into WPS branches”.
Conflation With Provenance – Custom WPS Processes
Similar to the efforts on Aviation-specific processing, 52°North developed processes for conflating two distinct data sets in the scope of the OWS 9 test bed. This article sheds some light on the details of the processes and the obstacles which need to be overcome.
Conflation is understood as a sophisticated process of unifying two or more separate data sets that share certain characteristics into one integrated all-encompassing dataset. In the process, different representations of a feature get folded into one feature. That target feature follows a particular model that is used to capture complementary information. In addition to the actual dataset conflation, the consistent capturing of provenance information was another goal of the project. The conflation processes developed represent a very customized and complex workflow. The test bed proved that the 52°North WPS framework fits perfectly into such use cases by providing an individual and flexible processing API.
The following pictures provides a high-level overview of the service architecture we established within the test bed.
Our WPS instance played a central role, acting as the main entry point for a client software. Conflation for OWS 9 has been split into two different phases:
- Geometry Conflation (performed via a customized 52°North WPS provided by GMU)
- Attribute Conflation (performed by the 52°North WPS with a Semantics Mediation Server back end provided by Envitia)
In addition to applying the actual conflation algorithms, another interesting part of this work involved the creation of provenance information. Since this architecture involves a high amount of automation in the processes, the results might not satisfy an expert user. In order to provide the applied process parameters to the user, the 52°North WPS created a fine grained set of provenance metadata, encoded as lineage types of ISO 19115-19139. The contents included information on the input data sets, the involved tool sets and details on each conflated feature (e.g. from which data set its attributes have been integrated).
Due to the fact that this architecture involves a complex chain of multiples processes and concepts, we cannot make the WPS branch source code deployable out of the box. Alternatively, you might be interested in taking a look at the OWS 9 demo instance. Please be aware that our service strictly requires the whole service chain to be available. In case of problems, please try at a later time.
This post will be updated as soon as the official OWS 9 demonstration page – featuring the 52°North Conflation with Provenance WPS – is online.
Previous Posts in this series:
Leave a Reply