Abstract
(English) | Deliverable D5.1 draws upon requirements identified in D1.1 and is derived from two running processes, namely the "Student Admission Process" and "SUAP". The "Student Admission Process" is a running process at FHNW and the "SUAP" is a running process at Marche Region. We illustrate the functionalities that will be implemented in a mock-up for the two real processes. The mock-up provides the golden threat throughout the document. The main artefact reported on in this deliverable is the Learn PAd ontology. Using ontology to represent information is widely accepted as it is machine understandable as well as cognitively adequate for humans. We reused an existing ontology namely ArchiMEO that was developed by FHNW. In order to cover Learn PAd specific requirements the ontology was enhanced regarding concepts and relations defined in the Learn PAd Conceptual Meta Model (LCMM) and the Learn PAd Platform Independent Meta Model (LPIMM) which are developed in WP3. We reassessed the ontology by posing competency questions in order to determine the formal terminology needed to meet the requirements. Since ontology development is considered an iterative process, the Learn PAd ontology represents a first version and will be continuously improved throughout the project. Since the ontology is supposed to support process execution and learning the ontological representations were related to non-ontological representations, particularly to the models of various kinds (e.g. a process modelled in BPMN). Therefore, we mapped the classes of LPIMM to concepts of the Learn PAd ontology. With this approach we can provide the semantics to infer the context of a process, which we illustrated using again the mock-up for the two processes. In addition to the mapping between models and the ontology we defined a mapping between models and their representations in XWiki. Furthermore, we introduced mechanisms for retrofitting modifications and annotations made in the wiki into their ontological representation.
|