The MELODIES Shared Platform

MELODIES will deliver a shared information technology platform that provides the services with all the “glue” necessary to understand and use Linked Data and environmental data.

The following presentation on this service was given at the MELODIES: Exploiting Open Data conference in October 2016. Slides are available here.

MELODIES will deliver a shared information technology platform that provides the services with all the “glue” necessary to understand and use Linked Data and environmental data; access data catalogues and endpoints through a variety of protocols; and to manage input, intermediate and output data. The shared platform will support the full life-cycle of the services by providing the following capabilities based on an underlying cloud computing infrastructure:

1. Development platform

The service developers will use the development platform (“sandbox”) to implement and test their applications on a stable environment identical to the final deployment environment, but safely separated from it.

2. Access to Linked Data and environmental data products

The service developer will define input datasets for the services, these datasets will then be available directly on the development platform. Selected shared datasets will be stored in a data staging area, and some data (together with other related information discussed above) will be stored in a Linked Data database, which enables sophisticated querying across diverse datasets.

3. Publishing of output data

The MELODIES project will publish project outputs as openly as the services in question allow. The project will therefore establish a data catalogue with a web service interface that complies with INSPIRE guidelines, building on previous developments by Terradue in GENESI-DEC and GEOWOW. Data, components and services will be published in the GEOSS Common Infrastructure for others to discover. MELODIES will therefore be among the first projects to demonstrate the interoperability between initiatives such as GEOSS, INSPIRE and the Open Geospatial Consortium standards and the “new world” of Linked Data.

4. Shared code libraries for visualization and mapping

All the services require some visualization capability: in some cases the capability is highly specific to the service in question (and hence is developed by the WP in question) but the MELODIES platform will provide some shared capability that can be exploited by all the services. This capability will include code libraries (Application Programming Interfaces) for generating visualizations from different sources of data, including Linked Data, based upon existing open-source software.

5. Application management

The service developer will benefit from application management to keep track of all changes done to his/her code and scripts. This approach guarantees that the environments can be restored and backup between working sessions.

6. Support for data processing

The technology platform will provide support for processing of various kinds, including a parallel processing environment that can be simulated in the service’s sandbox. Software toolboxes that provide the building blocks for data processing chains will be provided and, in some cases, developed within the project.

7. Flexibility of development and deployment

The MELODIES services require the flexibility to scale up and scale down their hardware requirements as the number of users increases and decreases in response to demand (e.g. in response to a flood event in the crisis management service). They require the ability to perform experiments and development in private “sandboxes”, but then easily deploy their final products in an operational system that shares the same computing environment. These requirements point to an infrastructure based upon cloud computing, in which users create “virtual machines” to their own specifications (disk size, processor speed and so on) that run on a shared physical infrastructure.