Foundational Definitions#

These are the stable teaching definitions for the beginner curriculum. Each term has one heading so other chapters can link to a predictable browser anchor. The definitions state how this project uses a term; they do not claim to settle every academic or professional debate.

Read From a missed collection to an ontology for the continuous example. Use the authoritative-source register to check the formal specifications, guidance, status and limits behind these definitions. The beginner glossary remains the broader quick-reference list.

Data#

Definition. Recorded representations used as inputs to a task: for example text, numbers, dates, identifiers, images or structured observations. Data is shaped by collection choices, formats and context; “raw” does not mean neutral or self-explanatory.

Boundary and example. A synthetic scheduled date and an observation that a container remains outside are data. Neither establishes why collection did not occur. Data quality is fitness for a declared purpose, not a universal score. See information, observation time, the on-ramp data section and the Government Data Quality Framework.

Information#

Definition. Data selected, organised or interpreted in context so that it answers or helps answer a question.

Boundary and example. Combining a collection date, waste type and place with the correct council's current schedule produces information relevant to whether a collection was expected. Clear presentation does not guarantee accuracy, freshness or applicability. Information and knowledge overlap in ordinary language; this curriculum uses knowledge when the justification and limits of a usable claim are explicit. See the on-ramp information section.

Knowledge#

Definition. A claim or model that can support understanding or action because its meaning, basis and limits are inspectable.

Boundary and example. “This synthetic event appears reportable under the selected local route observed on this date” can be knowledge when the rule, authority, evidence, derivation and scope are retained. Knowledge is not whatever an interface states fluently, and a schema-valid claim can still be wrong. See assertion, evidence, scope, DIKW and Evidence before fluency.

DIKW#

Definition. Data–Information–Knowledge–Wisdom (DIKW) is a teaching heuristic that asks how recorded values gain context and how conclusions support judgement.

Boundary and example. DIKW is contested, not a universal law or a required pipeline. Data is never wholly context-free, boundaries between information and knowledge vary, and more data does not automatically produce better judgement. This curriculum does not model “wisdom” as a machine-generated layer. See the full caution, data, information and knowledge.

Ontology#

Definition. An explicit model of kinds of things, identified properties and semantic relationships in a domain. An ontology may support inference, but its expressiveness depends on the language and profile used.

Boundary and example. A missed-collection ontology can distinguish expected collections, observations, reports, outcomes and complaints. It does not make four councils' local rules identical, and it is not merely a JSON object or a list of tags. Contrast taxonomy and controlled vocabulary. Continue with Vocabulary and ontology and the OWL 2 source boundary.

Taxonomy#

Definition. A classification that arranges concepts into a governed structure, commonly broader and narrower categories.

Boundary and example. A taxonomy may place household collections under rubbish, recycling and street services. It supports classification and browsing but need not define service episodes, evidence or formal inference. A taxonomy is not every navigation menu. See controlled vocabulary, ontology, SKOS and the GOV.UK taxonomy principles.

Controlled Vocabulary#

Definition. A governed set of permitted terms or codes, with maintained identifiers, labels, definitions and lifecycle decisions.

Boundary and example. A service-outcome vocabulary might permit acknowledged, rejected, return planned and unresolved. The control lies in governance and validation, not in alphabetical presentation. A controlled vocabulary can be flat; a taxonomy adds classification structure, and an ontology can add richer semantics. See Governed terms.

Identifier#

Definition. A value assigned to distinguish and refer to a thing within a declared identity system.

Boundary and example. A stable service-family ID should survive label changes. An identifier need not be a web address, and a human-readable label is not a reliable identifier. Local IDs need a declared namespace; global semantic identities use absolute IRIs in this profile. See URL, URI, IRI and Labels are not identifiers.

URL#

Definition. A Uniform Resource Locator is a web identifier that also provides a location or access mechanism.

Boundary and example. An HTTPS council page URL can identify where a reader retrieves current instructions. Availability, ownership and semantic identity are separate questions: a URL can redirect, disappear or serve new content. This curriculum uses credential-free HTTP(S) source URLs and checks their syntax and policy before generation. See URI, IRI, RFC 3986 and URLs.

URI#

Definition. A Uniform Resource Identifier is an identifier conforming to the generic URI syntax. It may identify by name, location or both.

Boundary and example. In the classic RFC distinction, a URL is a kind of URI associated with access. Not every URI is safely fetchable, and a syntactic URI does not prove that a resource exists. See URL, IRI, RFC 3986 and URI, URL and IRI.

IRI#

Definition. An Internationalized Resource Identifier extends the URI identifier model to a wider repertoire of characters while retaining defined mapping and syntax rules.

Boundary and example. RDF uses IRIs to identify resources and properties. This project's material semantic nodes and predicates use absolute IRIs plus validated local Explorer routes. An IRI can identify without being a page the Reader may fetch. See URI, identifier, RFC 3987 and RDF 1.1 Concepts.

Class#

Definition. A category whose members are things of a described kind.

Boundary and example. Public Service can be a class; one council's missed-collection route can be an instance. A class is not simply a display group or a tag, and membership should have declared semantics. In RDFS and OWL, class statements can participate in inference; validation is a separate layer. See Classes, instances and properties and RDFS.

Instance#

Definition. A particular thing represented as a member of one or more classes.

Boundary and example. A synthetic missed-collection report is an instance of a Report class. Two records with similar labels are not necessarily the same instance, and one real-world thing can have multiple records that require evidence-backed reconciliation. See class, identifier and Identity and similarity.

Property#

Definition. An identified characteristic or relationship used to make a statement about something.

Boundary and example. “Applies within jurisdiction” can be a property linking a service to a place. A local JSON key is not automatically an RDF property; semantic use requires a declared mapping to an IRI. When a property occupies the middle position of an RDF triple it is the predicate. See RDF and Classes, instances and properties.

Predicate#

Definition. The IRI in the middle position of an RDF triple, identifying the property that relates the subject to the object.

Boundary and example. In “report — submitted through — online channel,” submitted through is the predicate. Its human label can vary while its IRI keeps identity stable. An ordinary Markdown link must not be interpreted as a domain predicate. See triple, property, Directed and typed relationships and the link boundary.

Triple#

Definition. The basic RDF statement form: subject, predicate and object.

Boundary and example. A triple can state that one identified report was submitted through one identified channel. It does not, by itself, carry the claim's authority, evidence, derivation, status or observation time. This profile therefore keeps a direct triple synchronised with an evidence-bearing assertion. See RDF: the graph model and RDF 1.1 Concepts.

Assertion#

Definition. A statement represented in the knowledge product. In this profile, a material directed relationship is also represented as an evidence-bearing assertion record with stable identity and governance fields.

Boundary and example. The assertion that a report used a channel retains source and target identities, predicate, kind, labels, status, scope, authority, derivation, observation time, evidence and rights. An assertion is not true merely because it is present. See triple, evidence, status, scope and Assertion-level evidence.

Evidence#

Definition. Material that supports, challenges or helps assess a claim or whether a requirement is met.

Boundary and example. A dated official schedule can support what a council published; a synthetic observation can support what the example says was seen. Neither automatically supports every claim in the record. Evidence can be incomplete, conflicting or unavailable. It remains separate from authority, confidence and rights. See CCCEV and Evidence before fluency.

Authority#

Definition. The competence, mandate or recognised role that makes a source appropriate for a particular type of claim.

Boundary and example. A council is authoritative for its current local reporting route; a resident is the source for their observation; an ombudsman states its own complaint scope. Authority is claim-specific and does not prove freshness, availability or correctness. High confidence does not create authority. See Authority, derivation and freshness and Assertion status.

Provenance#

Definition. Information about the origin, custody and transformation of a claim, record or artefact.

Boundary and example. Provenance can record which source supplied a rule, when it was observed, which activity normalised it and which output was generated. Provenance supports assessment but does not itself decide trust or truth. See evidence, observation time, Provenance and PROV-O.

Rights#

Definition. The permissions, restrictions, duties and other legal or policy conditions governing access to and use of material.

Boundary and example. Rights can cover viewing, linking, copying, modification and redistribution. A source being public on the web does not make it open data. Rights are broader than a licence, and an open licence does not make a claim accurate or authoritative. See Licensing, DCMI Terms and OGL 3.0.

Licence#

Definition. An explicit grant of permission to use material under stated conditions.

Boundary and example. The Open Government Licence 3.0 permits broad reuse of covered information subject to its terms. It does not cover material that the provider excludes or third-party rights it does not own. A licence is one part of rights, not evidence of factual correctness. See OGL 3.0 and Markdown is content, not permission.

Confidence#

Definition. A documented estimate of uncertainty about an assertion under a stated method, scale and calibration.

Boundary and example. A model may assign 0.9 confidence to a proposed classification. That number does not make the assertion official, confer authority or measure relationship strength. A confidence value without its method is ambiguous. See authority, status, Confidence, strength and count and Evidence has several axes.

Status#

Definition. A governed label describing the state or origin of a record or assertion within a declared lifecycle or assertion scheme.

Boundary and example. Official, normalized, inferred, model-derived and editorial-example are assertion-origin statuses; draft, stable and deprecated are lifecycle statuses. The applicable scheme must be named. Status does not describe geographic or population coverage; that is scope. See Assertion status and Two independent labels.

Scope#

Definition. The declared boundary within which a claim, record, rule or evaluation applies.

Boundary and example. Scope may be one council, one jurisdiction, one observation date, a synthetic fixture or a defined source population. It is not a confidence score or an assertion origin. Explicit scope prevents a local example from appearing universal. See status, jurisdiction, observation time and Two independent labels.

Life Event#

Definition. A situation or change in a person's circumstances that can trigger needs, duties, rights or interactions with services.

Boundary and example. Birth, leaving school, moving home, learning to drive, bereavement and caring can be life events. People do not experience one universal sequence, and the same event can create different branches by jurisdiction and circumstance. A life event is not a department or a single transaction. See user need, service episode and CPSV-AP.

User Need#

Definition. What a person or organisation needs from a service to achieve the right outcome for them, grounded in research rather than assumed from a preferred solution.

Boundary and example. “Know whether my expected collection was missed and what to do next” is a need. “Receive an email” is usually a proposed solution or channel. Needs can span organisational boundaries. See outcome, channel, the GOV.UK guidance and Start with the person.

Public Service#

Definition. A capacity or activity through which a public authority or authorised provider helps people or organisations meet needs, exercise rights, fulfil duties or obtain an output under applicable rules.

Boundary and example. Reporting a missed collection can be modelled as a public service family with local routes. A service is not identical to its web page, form, channel, organisation or one person's episode. See service episode, channel, CPSV-AP and Open Referral UK.

Service Episode#

Definition. One bounded interaction or attempt involving a person and a service within a longer journey.

Boundary and example. Checking a schedule, submitting a report or making a complaint can each be an episode. An episode has time, channel, evidence and an outcome, including failure or unresolved state. It is not the public-service definition itself and need not be a successful transaction. See public service, channel, outcome and the branching example.

Requirement#

Definition. A condition, demand or information need that something or someone is expected to satisfy or address.

Boundary and example. A local reporting route may require a collection date and address context. A requirement should retain its source, jurisdiction, validity and assessment method. It is not automatically a legal duty, a software field or a decision rule. See evidence, CCCEV and JSON Schema.

Rule#

Definition. An explicit statement used to constrain, classify, route or decide what follows under stated conditions.

Boundary and example. “Do not accept a report before the locally stated time” may be represented as a sourced operational rule. A rule needs authority, jurisdiction and observation time; it must not be copied from another council because the labels look similar. A rule may implement or assess a requirement, but the two are not synonyms. See scope and CPSV-AP.

Channel#

Definition. A means through which a person or system accesses or communicates with a service, such as web, phone, post or in person.

Boundary and example. An online form is a channel, not the whole public service or user need. Channel availability, accessibility and failure should be recorded; digital access must not be assumed. See user need, public service, service episode, CPSV-AP and Accessibility.

Outcome#

Definition. The result or state reached for a person, service episode or process, including an unresolved or adverse result.

Boundary and example. Acknowledged, rejected, return collection completed, explained and unresolved are possible episode outcomes. An output such as a reference number is not automatically the desired user outcome. Never infer success from submission alone. See user need, service episode, redress and CPSV-AP.

Redress#

Definition. A route through which a person can seek review, correction, remedy or accountability after a problem, decision or unresolved outcome.

Boundary and example. A council complaint followed, where applicable, by an external ombudsman route is redress. It is not a guaranteed remedy, and scope, time limits and prerequisite stages belong to the responsible body. See outcome, jurisdiction, the missed-collection journey and Data and AI ethics.

Jurisdiction#

Definition. The geographic, legal or administrative area within which an authority, rule, service or decision applies.

Boundary and example. United Kingdom, England, Scotland, Wales, Northern Ireland and a local-authority area are different relevant jurisdictions. Place similarity does not establish legal applicability, and one UK-wide routing page does not prove a uniform process. See scope, rule, UK legislation data and the four local variants.

Observation Time#

Definition. The date or time when a source, fact or condition was observed for the purpose of the record.

Boundary and example. A council page observed on 10 August can support what the project saw then; it is not necessarily the page's publication date, the service-event time or the assertion-generation time. Recording precision matters: a year is not a day. See provenance, scope, Timeline and Assertion-level evidence.

Relationships Among The Terms#

The definitions are meant to be used together:

• a life event can trigger a user need;
• a public service can address part of that need through one or more channels and service episodes;
• requirements and rules apply within a jurisdiction and scope;
• episodes produce outcomes, including failure, and may lead to redress;
• an assertion can express the relationship with stable identifiers, a predicate, status, evidence, authority, provenance, rights and observation time; and
• a controlled vocabulary, taxonomy or ontology governs how selected parts of that model are named and connected.

That chain is a modelling aid, not a claim that every person experiences one linear journey.