ENVRI Knowledge Base

This page acts as a 'notebook' for testing the different kinds of queries, directed towards the ENVRI Knowledge Base (ENVRI KB), that are needed to support different types of activity. Sample queries can be submitted to the live knowledge base, and results displayed.

This service is part of the ENVIRplus project's "Data for Science" theme. It makes use of the ontologies for Open Information Linking for Environmental Science Research Infrastructures (OIL-E), which itself uses the concepts and vocabulary introduced by the ENVRI Reference Model.

For the ENVRI Knowledge Base, we identified four key knowledge capabilities that application of the semantic linking framework can facilitate:

  1. Surveying the technical landscape. The web of knowledge created by semantic linking should help us understand what technologies (including software, standards and vocabularies) are being used by environmental science RIs.
  2. Comparative solution analysis. It should be possible to compare solutions developed by environmental science RIs---specifically, given the knowledge of how technologies are used in their proper context, we should be able to compare developments in equivalent contexts.
  3. Gap analysis and component recommendation. Given a reference model for environmental science RIs, it should be possible to identify what is missing in the current development state of a given RI, and based on both that model and the solutions developed by other RIs, it should be possible to then make certain recommendations.
  4. Linked open research infrastructure. The web of knowledge created by semantic linking should itself be publicly accessible, machine-navigable, and provide a programmatic gateway to the services and data held by the RIs, including where available data provenance and resource catalogues.

It is in terms of these capabilities that we consider the different kinds of query that the Knowledge Base should support, whether those queries originate from users or underlie the operation of customised tools for RI modelling or visualisation.


The following namespace declarations are automatically prefixed to all queries submitted via this page:

If necessary, these declarations can be edited. Refreshing the page will reset them.

The ENVRI community

The following sample SPARQL queries can be submitted to the live knowledge base and their results viewed.

All query examples are fully editable, should you wish to experiment. All sample data is structured according to the OIL-E ontologies.

Please be aware that all results regarding specific research infrastructures are provisional, and should not be taken as being approved in any way by the infrastructures themselves at this point in development.

Lets start with ENVRIplus itself. ENVRIplus is a project that supports the ENVRI community to produce common solutions for problems shared by environmental science research infrastructure.

The following is a simple query to identify the various properties attached to any individual described in the ENVRI knowledge base, in this case the ENVRIplus project.

We can also portray the same information in graph form using a CONSTRUCT query.

Looking specifically at the ENVRI community, we can examine the specific RIs involved in ENVRI at present, clustered by domain.

We can then look at what the Knowledge Base holds regarding specific infrastructures, by adapting one of the earlier queries. For example, for EPOS:

The above query can be easily adapted to other RIs using the information from the graph of the ENVRI community.

Quality control in ENVRI RIs

The following query simply lists all the quality control behaviours already defined in the Knowledge Base based on initial surveys:

Given the URI of a quality control behaviour (or indeed any other behaviour), we can explore the characteristics of that behaviour in more detail. This query visualises PI quality control for data ingested by IAGOS:

This query visualises all quality control behaviours identified for IAGOS:

As can be seen, the amount of data gets quite dense quite quickly, highlighting the need for more tailored visualisations.

We can also look up quality control behaviours based on certain characteristics, e.g. the ones that operate for 'near real-time' datasets:

Exploring the ENVRI Reference Model

This query will provide a list of all ENVRI RM concepts or individuals that are necessarily related to the "add metadata" information action [wiki].

This query visualises the example SELECT query above as a graph.

We can expand the previous query to look at all information actions defined by the ENVRI RM (slightly intensive action; may be a delay in response).

Other queries

This query can retrieve the technologies associated with particular services:

The following query can be used to identify concepts (both classes and individuals) described using the term 'catalogue':

The next query can be used to identify the capabilities provided by a catalogue service:

While this query can identify the capabilities required by a catalogue service:

Other tools

A recent student project produced an ENVRI 'reference model visualiser'. The main idea behind this viewer was to provide recommendations of computational objects based on free-text requirements input by a user. It extracts nouns and verbs from the free-text inputs provided and searches the knowledge base for components that refer to those words in their label or description - a basic, initial example of combining natural language processing with information retrieval for the ENVRI Knowledge Base. The main queries supported are:

  1. What computational objects are described using a given keyword?
  2. What operational capabilities are provided by a given computational object?
  3. What operational capabilities are required from other components by a given computational object?