Skip to main content

Web-Didaktik Ontology

The Web-Didaktik is a formal ontology for knowledge organization created by Prof. Dr. Norbert Meder (University of Bielefeld, 2006). It defines three dimensions for classifying any piece of knowledge: what kind of knowledge it is, how it relates to other knowledge, and how it is presented. ORQO implements all three dimensions.

Web-Didaktik ontology structure

Dimension 1: Knowledge Types

Every piece of content is classified into one of four base classes, each subdivided into specific types. The base class captures the epistemological function — not the topic, but what role the content plays in understanding.

Orientation

Function: Helps the learner perceive and navigate a domain. Orientation knowledge creates awareness without yet enabling specific actions.

TypeDescriptionExample
HistoryHistorical events providing metacognitive context"The TCP/IP protocol was developed in the 1970s by DARPA..."
ScenarioEvent complexes from reality or realistic constructs"Consider a distributed system where three nodes must reach consensus..."
HypotheticalAssumed, unproven states"Suppose the database goes down during a write operation..."
StoryProblemScenario presented in narrative form"Alice runs a web shop. One day her payment gateway starts rejecting..."
VirtualWorld3D/spatial representations for explorationInteractive simulations, virtual labs
FactsObjective data and measurements"PostgreSQL 16 supports up to 32 TB per table"
SummaryBrief presentation of expanded material"In summary, the three consensus algorithms differ in..."
OverviewBroad outline of sequences or structures"This chapter covers: authentication, authorization, and session management"
ProblemIdentifies issues related to the topic"The challenge with eventual consistency is that reads may return stale data"
DidacticalGoalFormulates learning objectives"After this section, you will be able to configure a load balancer"

Explanation

Function: Provides reasons, arguments, and understanding. Explanation knowledge answers why things are the way they are and what things mean.

Why-Explanations — answering "why":

TypeDescriptionExample
ConclusionDerivation from premises"Since all requests go through the gateway, and the gateway logs all traffic, therefore all traffic is logged"
EvidenceChain of conclusions forming proofMathematical proofs, formal derivations
CausalCause-effect relationships"The outage was caused by a memory leak in the connection pool"
HistoryEvents as temporal consequences"The language evolved from C to C++ because developers needed object orientation"
FinalCauseJustifying means to ends"We use caching to reduce database load"

What-Explanations — answering "what":

TypeDescriptionExample
TheoremLogical assertions about facts"CAP theorem: a distributed system cannot simultaneously provide consistency, availability, and partition tolerance"
DescriptionRepresentation without derivation"A load balancer distributes incoming traffic across multiple servers"
Definition:MathematicalFormal mathematical determination"Let f: A → B be a bijective function..."
Definition:TermDetermination of essential characteristics"An API gateway is a server that acts as the single entry point for a group of microservices"
InterpretationDecoding documents or situations"The error message 'ECONNREFUSED' means the target machine actively refused the connection"

Other Explanation types: Case (Example, Counterexample), Argumentation, Assumption (Hypothesis, Idea), Reflection.

Action

Function: Knowledge about human actions, practices, and procedures — the "knowing how."

TypeDescriptionExample
Rule:IndividualNorms for individual action"Always validate user input before processing"
Rule:CooperativeNorms for cooperative action"All PRs require at least one review before merge"
Procedure:AdministrativeAdministrative procedures"To request access: submit form A, wait for approval, then..."
Procedure:InstructionManualEquipment/tool operation steps"Step 1: Install the CLI. Step 2: Run init. Step 3: Configure..."
Procedure:SocialNormSocial behavior regulations"During code review, critique the code, not the person"
ChecklistReminder lists for critical factors"Pre-deploy checklist: tests pass, migrations run, env vars set..."
PrincipleBasic propositions governing rules"Separation of concerns: each module should have one responsibility"
StrategyPurpose-rational alternative actions"For high-availability, deploy across multiple regions with failover"
LawCodified legal knowledgeGDPR articles, licensing terms
ExamplePractical instantiation of action knowledge"Here's how to implement the retry pattern in Python..."

Reference

Function: Points to where information lives — the "knowing where."

TypeDescriptionExample
Document:StatisticsNumerical data summariesPerformance benchmarks, usage reports
Document:ReportBibliographic information of reportsResearch papers, audit reports
Document:ProtocolFormal meeting summariesRFC documents, standards specifications
Document:EncyclopediaEntryEncyclopedia articlesWikipedia-style overviews
Document:HandbookSubject-specific reference worksAPI reference documentation
CommunicationCommunication materialsEmail threads, chat transcripts
CrossReference:AnnexSupplementary materialsAppendices, attached datasets
CrossReference:GlossaryIndexed word explanationsTerm glossaries, acronym lists

Dimension 2: Relations

Knowledge units don't exist in isolation. The Web-Didaktik defines 48 typed relations (25 canonical types with inverses) that capture how knowledge connects. These relations are what enable structured navigation.

Relation types diagram

Hierarchical Relations

Relations that express scope, generalization, and decomposition:

RelationInverseMeaning
GeneralizesSpecializesA is a more general form of B
PartOfWholeOfA is part of the whole B
ComponentOfComponentWholeOfA is a component within B
SubProcessOfParentProcessOfA is a sub-process of B
AspectOfAspectWholeOfA is one aspect of the broader B

Associative Relations (Asymmetric)

Relations with a clear directionality:

RelationInverseMeaning
BasisForUnderpinnedByA is the logical foundation for B
CauseOfEffectOfA causes B
PurposeOfMeansOfA is the purpose that B serves
ContextOfContextualizedByA frames the context for B
TimelineBeforeTimelineAfterA precedes B in time
ProcessOfProcessedByA is a process applied to B
EvaluationOfEvaluatedByA evaluates B
FunctionOfFunctionallyDeterminedByA is a function of B

Associative Relations (Symmetric)

Relations where direction doesn't matter:

RelationMeaning
SimilarA and B share significant characteristics
AlternativeA and B are interchangeable options
AnalogousA and B share structural parallels in different domains
OppositeA and B are in opposition or contrast
ComplementaryA and B complete each other
AffineA and B frequently co-occur

Didactic Relations

Relations specific to learning sequences:

RelationInverseMeaning
DidacticallyBeforeDidacticallyAfterA should be learned before B
BelongsToA belongs to concept container B
FormOfMaterialOfA is the form of content B

Dimension 3: Media Types

The third dimension classifies how knowledge is presented, not what it says:

CategoryTypes
PresentationText, Table, Diagram, Image, Video, Audio, Formula
InteractionSimulation, Exercise, Quiz, InteractiveScenario
CommunicationDiscussionPrompt, GroupTask, PeerReview, Tutoring

The typed relations enable 11 navigation directions — structured ways agents can traverse the knowledge graph:

DirectionWhat it followsUse case
deeperSpecializes, PartOf, ComponentOf, SubProcessOf, AspectOf"Tell me more detail about this"
broaderGeneralizes, WholeOf, ComponentWholeOf, ParentProcessOf"What's the bigger picture?"
whyEffectOf, UnderpinnedBy, DeterminedBy"Why is this the case?"
consequencesCauseOf, BasisFor, Determines"What follows from this?"
contextContextOf, ContextualizedBy"What's the surrounding context?"
relatedSimilar, Alternative, Analogous, Complementary, Affine"What's related to this?"
nextDidacticallyBefore, TimelineBefore"What comes after this?"
previousDidacticallyAfter, TimelineAfter"What came before this?"
examplesUnits of type Explanation:Case:Example or Action:Example"Show me a concrete example"
howUnits of type Action (any subtype)"How do I actually do this?"
sourcesUnits of type Reference (any subtype)"Where can I read more?"

This is what makes the Knowledge Engine fundamentally different from RAG. A standard retrieval system can only answer "what's similar?" — ORQO's agents can ask "go deeper," "why?", "show me how," or "what's the broader context?" and follow typed edges to precisely the right knowledge.

Validated by Research

The INTUITEL EU FP7 project (2014) demonstrated that just four of these relation types were sufficient for multiple distinct learning paths to emerge from a single knowledge domain. ORQO implements the full taxonomy of 48 relation types.