(Thinking has moved on since this Interim draft – please now see the current Information Model)

ELM Information Model for Learning Outcomes and Competences

Interim draft towards a CEN Workshop Agreement

By the InLOC team

2012-06-29

Foreword

The accompanying Interim Guidelines document gives background to, and explanation of this model.

This Interim Information Model is also available as the submitted PDF: 20120629_InLOC_Interim_Report_IM.pdf.

Issues arising from further discussion and consultation on the Interim documents are recorded at Interim Issues.

Contents

Introduction to the InLOC model

This CWA provides and defines the InLOC Information Model: InLOC – "Integrating Learning Outcomes and Competences" – models information defining intended learning outcomes and competences (LOCs). That information is important to personal, professional and vocational development, training and education, whether in the workplace or in school, vocational or higher education.

Representing such information, for the purposes of interoperability of learning (and other) technology, involves looking at how others have similarly represented such information effectively in the past, retaining what currently and appropriately works well, and adding what is needed. Two specifications that have been widely reused in this way are the Dublin Core Metadata Initiative's DCMI Metadata Terms and more recently the W3C's Semantic Web work on SKOS. The team brought their knowledge of existing specifications together, and assessed them against surveyed stakeholder requirements. It became apparent that existing specifications were not fit for this purpose, in different ways.

While some of the Dublin Core terms are envisaged as having literal ranges, other terms have ranges that are potentially structured – for example, DC terms dealing with authorship have the class "Agent" as their range. InLOC takes the innovative step of treating these different kinds of properties differently. The literal properties (including title and description) that are closely tied to their respective resources are used in the customary way. But for the more complex properties, which tend to be more loosely associated with the resource they are describing, InLOC introduces a new general purpose property class. Topic, level and credit value are often associated with intended learning outcomes, and are all commonly represented using more than one field, and InLOC uses this new class to represent them.

The resulting InLOC specification is both simple and highly expressive. While RDF has an extremely simple foundation, the complexity of the compound properties that InLOC needs to handle, and the extra information that needs to be given about relationships, invite richer treatment. Dublin Core is simple, but not directly designed for representing highly structured information. SKOS helps well in representing general-purpose hierarchical structures, but current practice in representing learning outcome and competence information needs more than taxonomic trees. Atom is another good example of a useful general purpose representation, but intended learning outcome and competence information has substantially different requirements. InLOC has taken up these challenges.

The InLOC team offer this new specification as a solution that is able to express the rich diversity of current learning outcome and competence information, so that systems using that information can interoperate effectively. Though it deals with information that is rich and complex, the InLOC specification is simple and flexible, and thus should be quite easy to implement and to work with. It would be especially gratifying were InLOC to come to be recognised as a useful new model for other areas involving structures with reusable components.

Scope of the InLOC Model

The InLOC Information Model is a model of information about intended learning outcomes or competences (LOCs), to enable its sharing across systems, sectors and domains. It distinguishes the main characteristics both of individual LOC definitions, and structures (often called "frameworks") of LOC information. LOC information is represented independently of its use in any particular context, target group, organisation or individual. This model is intended to facilitate linking and sharing LOC information across a wide range of related ICT systems, such as: university student information systems, recruitment systems, appraisal systems; e-portfolio and related systems used by individuals to manage their CVs or professional profiles; systems supporting the development of competence, whether owned by an individual or a company; and relevant social networks, whether private ones operated by educational institutions or businesses, or public ones open to all on the Internet. There are many and varied uses of this information, and they could be categorised in several ways. The central point is simply that they all use, manage or refer to information about learning, development and competence.

InLOC does not set out any LOC concept definitions or structures, but rather, it provides a common format in which those definitions and structures may be represented, so that ICT systems can handle the information in standardised ways, more efficiently and effectively than would otherwise be possible.

InLOC's scope is focused on representing the information just about LOC definitions and structures, independently of their being applied to particular people. Through this clear scoping, InLOC structured information can be referred to not only in the context of individuals claiming to have attained learning outcomes or competences, but also from representations of course ("learning opportunity") information, and representations of information about resources for learning education or training, irrespective of the media used by those resources.

InLOC expressly does not attempt to represent any information about individuals, about courses or about other resources for learning, in the belief that the most effective and flexible way of representing information across the whole learning technology sector is to work towards standardising the representation of limited, well-defined, self-contained areas of information, together with simple ways of cross referencing between these separate representations. It is an approach that has been described in the phrase "small pieces, loosely joined".

Terms, definitions and abbreviations

For the purposes of this documentation, the following terms and definitions apply.

competence

the proven ability to use knowledge, skills and personal, social and/or methodological abilities, in work or study situations and in professional and personal development (EQF)

learning outcome

statement of what a learner is expected to know, understand, or be able to do after successful completion of a process of learning (ECTS Users Guide)

NOTE: The full proper term is "intended learning outcome", and unless it is clear from the context that the reference is to the actual outcome of learning in a person, "learning outcome" should be understood to mean "intended learning outcome".

LOC

learning outcome or competence

LOCs

learning outcomes and/or competences

NOTE: The intention of the term "LOC" is to play a role similar to the recently popularised term "competency", covering "knowledge", "ability", and other such concepts. The advantage of the term "LOC" is that, being artificial, it can be designated explicitly to apply to learning, education and training, to recruitment and employment, and to personal and professional development.

The InLOC Model specification

The InLOC Model is built on the basis of:

• distinguishing assessable LOC definitions from the LOC structures that contain and relate them;
• giving a common structure to compound properties of LOC definitions or structures;
• giving a common structure, including a number, to relations between LOC definitions and structures.

Many examples of existing definitions, frameworks and structures, related to learning outcomes and competences can be represented according to this model, with minimal levels of ambiguity.

InLOC Model Core Concepts

LOC definition

A LOC definition (class name: LOCdefinition) instance expresses a concept of a learning outcome or competence, against which a person (or, less often, a group or organisation) can be assessed. It is through being defined that people can understand what the concept refers to, and thus agree on the assessment of a person against that concept. Definitions of the same concept are equivalent if all assessment outcomes are the same whichever of the definitions is used. Assessment does not have to be grading. A qualitative assessment based on a collection of related presented evidence is still an assessment.

If a description describes something that someone is able to do (a competence), or something that someone may have learned (a learning outcome), it could belong to a LOC definition. If the described concept is not related to individual knowledge, ability, or personal characteristics, or if there is no way of making a claim to the concept and supporting it with evidence, then the concept does not fit as a LOC definition.

LOC definitions directly and logically contain only their Simple atomic properties, which is the information (or metadata) essential either to the defining the meaning of the described LOC concept, or to its immediate use or reuse. Compound properties of a LOC definition are represented using the LOCproperty class, but as these other properties attributed may vary according to different authorities, the LOC properties belong not to the LOC definition, but instead to a containing LOC structure.

LOC structure

A LOC structure (class name: LOCstructure) instance serves to define a set of LOC definitions that are included in that structure, to define the relations that exist between those LOC definitions, and to define LOC properties of the same LOC definitions. LOC definitions are logically included in a LOC structure through explicit LOC relations, rather than directly by means of containment through the presentation or format. It is immaterial whether the representation of LOC definitions is sequentially included in the LOC structure or not. This means that LOC definitions from elsewhere can be as easily included as ones defined along with the LOC structure.

A LOC structure instance will have properties of its own. Simple properties are represented as Simple atomic properties of the LOC structure, and compound properties as LOC property instances, with the LOC structure identified as the subject of the LOC property instance.

To allow reuse of whole frameworks, lesser LOC structures can be represented as part of greater LOC structures.

A LOC structure may represent what is called a "framework" or "occupational standard" in existing examples.

LOC property

LOC property (class name: LOCproperty) instances are the compound properties of LOC concepts. Whereas Simple atomic properties have a single value, LOCproperty instances have their own internal structure. The LOCproperty class structure helps to ensure that datatypes are predictable and information is processed easily. A LOC definition or a LOC structure may be the subject of a LOC property, but the LOC properties logically belong to the innermost containing LOC structure, not a LOC definition, even if their presented position is inside the representation of a LOC definition.

LOC relation

LOC relation (class name: LOCrelation) instances, representing relationships between instances of LOCdefinition or LOCstructure, comprise a major part of the content of a LOC structure. Because LOCrelation instances have no date and authorship metadata of their own, they inherit these and other properties from their containing LOCstructure. They must therefore appear contained within the LOC structure they belong to. Even if their presentation happens to be within a LOCdefinition instance, they are not regarded as logically contained by it. LOC relations cover relations between LOC definitions, between LOC structures, and either way between LOC definitions and LOC structures.

InLOC Model diagram

Figure 1 illustrates the InLOC Model using some UML conventions.

Each box in the diagram relates to a Class. No cardinality is specified for any association. Lines with an unfilled triangle shape indicate a subclass relationship. Lines with a filled diamond shape represent an association with a composition relationship, where the associated parts have no independent existence.

InLOC Resources

Note on the use of the prefix "inloc:"

The prefix "inloc:" is used both as a namespace and a CURIE prefix. The current intention is for "inloc:" to stand for and be substituable by the URI "http://purl.org/net/inloc/", but if a better option is provided by CEN before the finalisation of any CWA or EN, this will be considered.

Classes

The following classes are defined for InLOC.

Attribute Properties

Simple atomic properties of the LOCdefinition and LOCstructure classes

Simple atomic properties that are the same as the corresponding ones from DCMI Metadata Terms

Simple atomic properties defined within InLOC

Properties shared between the LOCrelation and LOCproperty classes

Properties of the LOCrelation class

Properties of the LOCproperty class

Related documents

• The Interim Guidelines presents the accompanying guidelines and request for comments.
• The XML InLOC project web page links to some very early drafts of example XML, together with an XSD.
• Example analyses are available from the Stakeholders InLOC project web page.