• Distributed e-Learning

This is one of the possible Use Cases.

1. Abstract

This use case shows several applications of rules for distributed e-learning. Learners (or, more specifically, their agents) contact multiple learning material providers to retrieve learning material or compute learning paths (workflows over learning material), with the goal to fill a knowledge gap between the current knowledge and a learning goal, taking user profile (language, learning method, ...) and also privacy considerations into account. Meta-data for learning material and personal profiles are expressed wrt. some ontologies, incl. a domain ontology for learning topics.

2. Status

• Originally proposed by MichaelSintek (DFKI GmbH) at F2F1 meeting.

• This use case is based on experiences (incl. prototypical implementation) of the ELENA project, an European research initiative funded under the IST program(me) of the European Commission from September 2002 to May 2005.

3. Links to Related Use Cases

• Enterprise_Information_Integration is (to some extent) a more general description of our use case

• Operationally_Equivalent_Translations are required for the transformation/generation of rules in our use case

• Ontology_Mapping_with_OWL_and_Rules is also needed for our use case if OWL ontologies are used to describe learning material

• Scoped_negation,_Encapsulation is obviously needed for distributed inferencing

• Publication of semantics (e.g. SKOS, RDFS) is needed for distributed material providers to announce their capabilities

• Automatically_generated_rules: our use case requires the generation of rules (where the generation is performed by other rules)

4. Relationship to OWL/RDF (and XML) Compatibility

• The ontologies (and schemas) for learning resources at the learning material providers are formulated in different ontology languages (RDFS, OWL-Lite/DL, XML Schema, ...), therefore compatibility of RIF to these languages must be provided (even in a single inference for rule/ontology transformation/mapping, see below).

5. Examples of Rule Platforms Supporting this Use Case

• This use case has been prototypically implemented with the rule language TRIPLE.

6. Benefits of Interchange

• rules (and queries) are sent to learning material providers which support different rule/query languages and ontologies/ontology languages
• rules themselves have to be preprocessed (for ontology mapping, restricted rule language support by some providers, removing parts revealing privacy-relevant information, etc.)

7. Requirements on the RIF

• RIF must support the representation/tagging of various rule dialects/semantics (in a way that rule engines can announce their capabilities in these terms and compatibility between a rule set and a rule engine can easily be determined), i.e., we probably need an ontology of RIF semantics (formally represented in a standard ontology language)

• RIF rules should be appropriate for distributed inferencing
• RIF rules should be represented in a way that other RIF rules can transform them, e.g., for ontology mapping, transforming rules (e.g., for distributed inferencing), etc.

8. Breakdown

8.1. Actors and their Goals

• Learner - wants to retrieve learning material or learning paths to reach some learning goal
• Learner Agent - agent acting on behalf of Learner; has personal profile containing already learned material and personal preferences
• Learning Material Provider 1 .. n - offers learning material (electronic, real courses, ...) some of which involve payment; material is organized wrt. local ontologies

8.2. Main Sequence

1. Learner instructs Learner Agent to retrieve learning material/paths by specifying a learning goal
2. Learner Agent creates queries/rule set to compute learning material/paths based on the learning goal and the personal profile
3. Learner Agent contacts appropriate Learning Material Providers and retrieves information on them, including the capabilities of their query/rule engine
4. Learner Agent analyzes queries/rule set wrt. the capablities of the Learning Material Providers and privacy considerations and transforms them accordingly (including ontology mapping, simplification, etc.), resulting (in general) in an execution plan (workflow)
5. Learner Agent sends transformed queries/rules to Learning Material Providers
6. Learning Material Providers execute queries/rules and send results back to Learner Agent
7. Learner Agent sends additional queries/rules to some Learning Material Providers according to execution plan
8. Learner Agent combines results and presents them to the Learner

9. Narratives

(replace Learner and Learning Material Provider i by names of your choice in Main Sequence and drop enumeration )

10. Commentary

• inference mechanism (RIF engine) in Learner Agent must be able to handle multiple ontology languages within one inference (since Learning Material Providers use various ontology languages)
• to allow RIF rules to transform other RIF rules, one solution is to have an RDF/OWL-based syntax for RIF; RIF is required to work on RDF/OWL (instances)
• this use case will usually also involve various kinds of impreciseness (the user wants, at the end, a list of learning material/paths ordered according to his preferences/costs/...), but to keep this use case simple, this aspect was skipped