ESA has been performing several activities during recent years to explore the applications of Semantic Technologies to bridge the gap between users who need meaningful information derived from EO systems, and the enormous amount of data available at ESA archives. Following up the results of these activities, ESA wants to apply these technologies on Marine and Coastal Applications in the Mediterranean Area, and compare its results with the outcomes of similar activities that are being performed by the Mississippi State University (MSU) in the Gulf of Mexico. Accordingly, ESA has put in place a cooperation agreement with MSU in order to exchange information, identify common approaches and compare results. The present project is aimed at supporting this cooperation and to produce results in line with current ESA approach for ontology implementation. Objectives The SDD project general objective is summarised as follows: The "Semantics Driven Framework for Resource and Knowledge Discovery" (SDD) project general objective is to develop, test and verify with MSU and other partners (e.g.: Experts, EUSC, EEA) • an ontology based implementation • supporting the identification and use of • a number of marine and coastal applications in the Mediterranean area through seamless, semantic access to the information they provide from EO images and other sources. Distilling the project title, "Resource and Knowledge Discovery" represents the aimed functional capability of SDD (WHAT the SDD system shall do), whereas "Semantics Driven Framework" represents the solution approach (HOW to do it), particularly referring to the emerging field of ontologies and its technologic implementation following W3C's web standards on semantic technologies. Translating this objective into a system requirements engineering vocabulary, we get the following top level functional and architectural requirements: • The SDD system shall be able to support the user in identifying and using relevant EO information resources • The SDD system shall be based on a semantic driven framework based on an ontology based implementation. The SDD project is initially focused on the maritime domain, nevertheless the results and lessons learnt during the work are expected to be potentially extrapolated (in subsequent phases) to other domains of applications in which EO data can provide an added value. The project is initially targeted towards the Experts and potential end users in the context of: • The European Environmental Agency (EEA) • The European Union Satellite Center (EUSC) • The Integrated Coastal Zone Management initiative (ICZM) • The ESA GMES Service Element MarCoast network (MARCOAST) The SDD project has been structured in three main parts: • Applications, Services and Knowledge Discovery. There is an Experts Survey and Analysis that focuses on people and organisations. It provides a survey and analysis of organisations/projects and key personnel who hold suitable knowledge and expertise and could be approached to support the work of this project. They can represent organisations involved in service provision, in regulation, or user activities. The section provides a table of experts and proposed level of involvement with the project. There is also an Applications Analysis section that focuses on applications, on the systems that have been assembled to provide the infrastructure and supply chain necessary to deliver the services. This refers specifically to the data processing chain. • Classification and Terminology Knowledgebasesa. The main goal of this section is the creation of the Classification KB (i.e., taxonomies), the Terminology KB (i.e., thesaurus) and their verification and validation with Experts. • Demonstrator Implementation. In this section the main goal is the selection of the final Applications, analysis of Application Implementation Requirements, detailed Definition of Proposed Demonstrations, negotiation of Agreements with Application, Data & Sevice Providers for Integration and Applications Integration & Tests. Architecture The vision of deployment of this model is illustrated in the figure below. In the figure, the blocks enhanced with red-border represent the focus of the SDD project in a wide context • The Classification Knowledge bases: Can be common, i.e., multi-domain. Can be created from Knowledge Discovery tools.. • The Portfolio Databases: are specific for each system or service. Should be implemented for EO using (linking) the shared Terminology Database (in order to minimise confusions and improve automated checks, for example for service chaining) The Classification KB usually takes the form of taxonomies, i.e., hierarchical schemas of service and products. The Portfolio databases correspond mainly to operational catalogues of services and products, i.e., operational metadata systems. Several international standards do already address this part (e.g., ISO, OGC), and therefore their study and integration will be object of the SDD work. The terminology management refers to shared thesaurus in use (or potentially in use) in the domains of application. They can be also common, and represent a supporting mechanism to user in correctly identifying the relevant resources. In summary, the SDD project addresses the above points by aiming at: • Creating the necessary content for the Classification Knowledgebase for a set of domains, and the shared Terminology and the Portfolio Databases for selected services (providing EO or non-EO products) • Creating and testing demonstrators for selected marine and coastal applications using as much as possible existing tools like Knowledge-based Information Mining (KIM), Knowledge-centred Earth Observation (KEO) and Service Support Environment (SSE) • Comparing the approach and the results with MSU • Defining requirements, suggestions and paths for future activities How it will work After identifying, cataloguing and analysing the key marine applications, we focus on the Evaluate Demonstrator, that basically tries to demonstrate that the outcomes from the SDD project prove the advantages of applying web semantics on several application domains engaging experts within the MARCOAST context. Several demonstrators (based on the description of services and products from different domains) are planned to be deployed and tested, to prove the feasibility of making discovery processes more powerful: • Water Quality Monitoring • Fisheries Support Services (HABs) • Ocean State Monitoring • Watershed Monitoring • NRT Wind and Wave Monitoring • Coastal Zone land Use / Land Cover These demonstrators will rely on the KEO Resource Browser. It is intended that they use datasets provided by several MARCOAST partners, in order to help the evaluation process of the SDD semantic added value.
