Glossary
Working explanations for scientists entering semantic engineering. Definitions are contextual; individual communities and standards can draw boundaries differently. These notes are not new ontology axioms or approved concept mappings. See key distinctions, resource landscape and human review workflow.
Source orientation: SKOS, SOSA/SSN, OBOE, I-ADOPT recommendations, Darwin Core, PROV-O, QUDT and CF conventions. The prose below is project explanation; consult the cited resource for normative definitions.
Find a term
Benchmark profiles: explicit ontology use, ontology potential, semantic diversity, ontology leverage, breadth, depth, composability, complementarity, ontology accumulation.
Semantic infrastructure: ontology, vocabulary, controlled vocabulary, thesaurus, classification, taxonomy, knowledge graph, semantic model, conceptual model, schema, standard, concept, class, instance, identifier, URI / IRI, label, synonym, definition, axiom, relation, subclass / is-a, part-of, member-of, broader / narrower, equivalent, alignment, mapping, crosswalk, semantic interoperability, upper ontology, domain ontology, application ontology, ontology import, ontology profile, semantic composition, entity-quality model, I-ADOPT decomposition, SKOS.
Biological organization: biological entity, organism, taxon, taxon name, taxon concept, taxonomic rank, population, community, assemblage, guild, functional group, ecosystem, biome, habitat, environmental system, environmental feature, environmental material.
Biological characteristics: trait, attribute, phenotype, quality, behavior, biotic interaction, trophic interaction, property hint.
Observation and measurement: observation, measurement, observed property, feature of interest, object of interest, matrix, context object, sample, specimen, occurrence, event, inventory, survey, sampling, protocol, method, sensor, platform.
Scale and support: spatial extent, spatial resolution, spatial support, footprint, CRS, location, temporal extent, temporal resolution, temporal support, timestamp, instant, interval, scale, biological scale, ecological organizational level, grain, extent, support, aggregation, upscaling, downscaling, cross-scale transformation.
Quantitative semantics: variable, quantity, quantity kind, unit, dimensionless, calibration, quality control, missing value, uncertainty, confidence.
Provenance and harmonization: provenance, lineage, entity/activity/agent, derivation, transformation, harmonization, standardization, normalization, semantic equivalence, numeric convertibility, scientific comparability, AI-ready data, representation, embedding, core harmonization, track-specific harmonization, evidence, source-documented evidence, observed evidence, deterministically derived evidence, model judgment, human decision, needs evidence, semantic gap.
Semantic infrastructure
ontology
A formally structured account of entities, relations and constraints in a domain.
Why it matters: Its commitments can support logical inference; a list of labels alone cannot.
Related: controlled vocabulary, knowledge graph.
vocabulary
A set of terms used to describe something.
Why it matters: A vocabulary may supply identifiers and definitions without formal axioms.
Related: controlled vocabulary, ontology.
controlled vocabulary
A managed set of permitted terms with documented meanings.
Why it matters: Controlled labels reduce variation but do not prove that measurements are comparable.
thesaurus
A managed vocabulary with broader, narrower and associative links, often multilingual.
Why it matters: Thesaurus broader is not automatically logical subclass.
Related: broader / narrower, SKOS.
classification
A scheme assigning things to named categories.
Why it matters: Categories can serve a reporting purpose without describing every entity relation.
taxonomy
A system for classifying organisms, or more generally a classification hierarchy.
Why it matters: Biological taxonomies are versioned scientific opinions, not one immutable organism ladder.
Related: taxon, taxon concept.
knowledge graph
An interconnected set of assertions about identified things.
Why it matters: A graph can use an ontology, but instance assertions and ontology definitions are different products.
semantic model
A structured description of meanings and relationships used by an application or community.
Why it matters: It may be conceptual rather than an executable formal ontology.
Related: conceptual model, schema.
conceptual model
A high-level account of important entities and how they relate.
Why it matters: A diagram does not necessarily supply formal predicates or inferencing rules.
Related: semantic model, axiom.
schema
A structural contract for fields, types and constraints in a data format.
Why it matters: Passing a schema validates structure, not the scientific truth of a record.
Related: standard, semantic model.
standard
A community-agreed specification, such as an exchange format, checklist or terminology.
Why it matters: A standard need not be an ontology and compliance need not imply harmonization.
Related: ontology, standardization.
concept
An identified meaning used to organize or describe knowledge.
Why it matters: One label can express different concepts and one concept can have several labels.
Related: identifier, label.
class
A category whose instances satisfy a specified meaning or definition.
Why it matters: A class of organisms is distinct from a particular observed organism.
Related: instance, subclass / is-a.
instance
A particular thing represented as belonging to a class.
Why it matters: A particular tree is not the class tree or a taxon name.
identifier
A persistent reference used to distinguish an entity or concept.
Why it matters: Stable identity survives label changes; version and authority still matter.
URI / IRI
A globally structured identifier; IRIs permit a wider character repertoire than URIs.
Why it matters: An HTTP identifier may resolve to documentation, but resolution alone does not establish semantic correctness.
Related: identifier, definition.
label
Human-readable text associated with an identifier.
Why it matters: Identical labels do not establish equivalence; different labels do not establish difference.
synonym
An alternative designation associated with a term under a stated synonym policy.
Why it matters: Exact, related and historical synonyms have different implications.
definition
A statement specifying what a concept means.
Why it matters: Compare definitions, entity types, methods and support rather than matching labels alone.
Related: axiom, semantic equivalence.
axiom
A formal assertion or constraint used in an ontology's logical model.
Why it matters: Reasoner conclusions depend on these commitments and their consistent application.
Related: relation, subclass / is-a.
relation
A specifically defined connection between identified things.
Why it matters: Part-of, member-of, subclass and broader are different relations.
subclass / is-a
A class-inclusion relation: every instance of the narrower class is an instance of the broader class.
Why it matters: This is not physical containment or population membership.
Related: part-of, broader / narrower.
part-of
A relation between a whole and a constituent, under a specified mereological definition.
Why it matters: A leaf can be part of a plant; that does not make leaf a subclass of plant.
member-of
A relation linking an individual to a collection or group.
Why it matters: An organism can belong to a population without being a physical anatomical part of it.
Related: population, part-of.
broader / narrower
Concept-scheme links expressing more general and more specific meanings.
Why it matters: In SKOS these links do not by themselves assert OWL class inclusion.
Related: subclass / is-a, SKOS.
equivalent
The same relevant meaning under an explicit semantic regime and scientific context.
Why it matters: A strong equivalence claim requires evidence, versioning and appropriate review.
Related: semantic equivalence, numeric convertibility.
alignment
A documented set of proposed relationships between semantic systems.
Why it matters: Alignment can include broader, narrower or related links, not just equivalence.
mapping
An evidenced assertion relating a source concept to a target concept.
Why it matters: It must preserve the source identity and state relationship, uncertainty and review status.
Related: alignment, harmonization.
crosswalk
A maintained collection of mappings between schemas, vocabularies or classifications.
Why it matters: A crosswalk may be lossy, contextual and version-specific.
Related: mapping, provenance.
semantic interoperability
The ability to exchange and interpret information while retaining relevant meaning.
Why it matters: Readable files or matching field names are insufficient.
Related: harmonization, schema.
upper ontology
A model of very broad categories such as objects, processes and qualities.
Why it matters: It provides foundational commitments rather than all domain terms.
Related: domain ontology, application ontology.
domain ontology
An ontology describing a scientific or practical subject area.
Why it matters: Its intended scope is a strength and a boundary, not a universal failure.
Related: upper ontology, semantic composition.
application ontology
An ontology assembled or specialized for a defined task or application.
Why it matters: Its useful assumptions may be inappropriate elsewhere.
Related: domain ontology, ontology profile.
ontology import
An explicit mechanism for incorporating another ontology's axioms or modules.
Why it matters: A citation or similar coverage is not evidence of an actual import.
Related: axiom, semantic composition.
ontology profile
A documented selection of semantic resources, patterns and constraints for a purpose.
Why it matters: A profile composes community work without replacing every source ontology.
Related: semantic composition, mapping.
semantic composition
Building a meaning from explicitly related concepts and constraints supplied by several resources.
Why it matters: Composition requires compatible relations and evidence; juxtaposed labels are insufficient.
Related: entity-quality model, I-ADOPT decomposition.
entity-quality model
A pattern expressing a quality of a specified entity, potentially with further context.
Why it matters: A quality alone cannot specify which entity, method or support was measured.
I-ADOPT decomposition
A variable description using property, object of interest, optional matrix/context and constraints.
Why it matters: The framework does not supply every domain concept, unit, method or observation model.
Related: object of interest, matrix.
SKOS
A W3C framework for representing concept schemes, labels and semantic links.
Why it matters: It can represent a thesaurus without imposing all the axioms of a domain ontology.
Related: thesaurus, broader / narrower.
Biological organization
biological entity
A biological thing or collection identified under a stated scientific model.
Why it matters: An organism, taxon concept, population and anatomical part are not interchangeable entities.
organism
An individual living entity identified at an appropriate level for the study.
Why it matters: A tree observation concerns an individual, not automatically its species or the plot community.
Related: taxon, population.
taxon
A taxonomic grouping recognized under a classification or taxon concept.
Why it matters: A taxon is distinct from its name and from an ecological population.
Related: taxon name, taxon concept.
taxon name
A nomenclatural designation applied under a naming code.
Why it matters: The same name can have different circumscriptions in different taxonomic treatments.
Related: taxon concept, identifier.
taxon concept
A taxon's circumscription according to an identified treatment, authority or version.
Why it matters: A name match without the treatment may hide scientific differences.
Related: taxon name, mapping.
taxonomic rank
A category such as species, genus or family in a taxonomic system.
Why it matters: Rank is not an organism's size, ecological level or observation support.
Related: taxonomy, biological scale.
population
A scientifically specified group of organisms, often sharing taxon, place and time criteria.
Why it matters: Population membership and boundaries must be stated; they cannot be inferred from a raster footprint.
Related: community, member-of.
community
A study-defined set of populations or organisms considered together in an ecological context.
Why it matters: Composition depends on inclusion criteria, location, time and detection methods.
Related: population, assemblage.
assemblage
A set of organisms grouped for a study, often within a selected taxonomic group.
Why it matters: Usage differs across communities; state the operational definition.
guild
A grouping based on comparable resource use, under a stated ecological definition.
Why it matters: Resource-use categories are not necessarily taxonomic or phylogenetic groups.
Related: functional group, trophic interaction.
functional group
Organisms grouped by selected ecological functions or traits.
Why it matters: Grouping depends on the chosen function and cannot replace source taxa silently.
ecosystem
A specified ecological system involving organisms and their physical environment and interactions.
Why it matters: Its boundaries and measured properties need explicit study definitions.
Related: community, environmental system.
biome
A broad ecological regional grouping characterized by major environmental and biological patterns.
Why it matters: Biome classifications differ and are not a universal spatial support scale.
Related: ecosystem, classification.
habitat
The environmental setting used by an organism or group under a stated definition.
Why it matters: A habitat category is not the same as a measured environmental material or a polygon boundary.
Related: environmental system, environmental feature.
environmental system
An environmental whole described as a system under a particular semantic model.
Why it matters: System identity and boundaries must be distinguished from a sampled material.
Related: environmental feature, environmental material.
environmental feature
An identifiable environmental feature, such as a landform, in a resource's terminology.
Why it matters: A feature may contextualize an observation without being its sampled matrix.
Related: environmental system, matrix.
environmental material
A substance or material in the environment, such as soil or water.
Why it matters: The material containing a measured chemical differs from the surrounding ecosystem.
Related: matrix, context object.
Biological characteristics
trait
A characteristic of a biological entity under a specified trait definition.
Why it matters: A trait is not the stored variable name or the value of one measurement.
Related: variable, measurement.
attribute
A characteristic used to describe an entity or record.
Why it matters: Data-field attributes and biological attributes can have different meanings.
Related: trait, property hint.
phenotype
An observable characteristic or set of characteristics of an organism in a stated context.
Why it matters: Phenotype descriptions depend on entity and conditions; they are not simply genotype labels.
Related: quality, entity-quality model.
quality
A characteristic associated with an entity, as defined by the semantic resource in use.
Why it matters: Ontology quality differs from data-quality acceptance or uncertainty.
Related: trait, quality control.
behavior
An action or process involving an organism under an explicit behavioral definition.
Why it matters: Behavioral categories do not automatically imply ecological interactions or comparable observation protocols.
Related: biotic interaction, phenotype.
biotic interaction
A specified relationship or process involving living entities.
Why it matters: Direction, participants, context and source evidence matter.
Related: trophic interaction, relation.
trophic interaction
An interaction concerning consumption or resource transfer under a stated model.
Why it matters: Trophic category assignments may vary with life stage, context and evidence.
Related: guild, biotic interaction.
property hint
A recorded clue about which property a source describes.
Why it matters: A hint is not an accepted canonical property or approved mapping.
Related: observed property, model judgment.
Observation and measurement
observation
An act or record of determining something about a feature or entity.
Why it matters: SOSA and OBOE organize observations and measurements differently; preserve the source model.
Related: measurement, feature of interest.
measurement
A determination of a property's value using a stated procedure and context.
Why it matters: A number without entity, property, unit, method and support is not a fully interpretable measurement.
Related: variable, uncertainty.
observed property
The characteristic targeted by an observation.
Why it matters: Two sensors at the same location may observe different properties.
Related: feature of interest, quantity kind.
feature of interest
The entity or feature about which an observation determines a property.
Why it matters: In an observation model it can differ from the sample used to access it.
Related: object of interest, sample.
object of interest
The entity whose property defines a variable in I-ADOPT.
Why it matters: Its semantic role is variable-specific, not a fixed label for an entire dataset.
Related: matrix, I-ADOPT decomposition.
matrix
The entity or material in which the object of interest is situated for a variable description.
Why it matters: A chemical concentration in water is not the same variable as concentration in soil.
Related: environmental material, object of interest.
context object
An additional entity supplying context to a variable description.
Why it matters: Context helps distinguish otherwise similar variables without replacing the object of interest.
Related: matrix, I-ADOPT decomposition.
sample
A selected entity, portion or proxy used to learn about a feature of interest.
Why it matters: A material sample and an observation's spatial sampling support need separate descriptions.
Related: specimen, spatial support.
specimen
A particular collected or preserved biological entity or material under a collection model.
Why it matters: A specimen can support many observations; it is not the observation event itself.
Related: sample, observation.
occurrence
Evidence of an organism or taxon at a place and time under a recording convention.
Why it matters: An occurrence record does not necessarily estimate abundance or sampling completeness.
event
A contextual occurrence or activity grouping observations under a specified data model.
Why it matters: Event identity does not by itself define survey design or detection probability.
inventory
A survey-derived account of entities observed within an explicit scope.
Why it matters: Inventory completeness requires effort and detection context, not just a species list.
survey
A planned observation effort using a sampling design and scope.
Why it matters: Different survey designs can make similar-looking results incomparable.
sampling
The selection or creation of samples according to a design.
Why it matters: Sampling design links observations to inference targets and must survive harmonization.
protocol
A specified procedure for carrying out an observation or experiment.
Why it matters: Protocol identity and version can alter measurement meaning even when units match.
Related: method, measurement.
method
The documented approach used to obtain or derive a result.
Why it matters: A method label may be less specific than a reproducible protocol.
Related: protocol, calibration.
sensor
An instrument or system used to determine a property.
Why it matters: A sensor is distinct from the platform carrying it and the property it measures.
Related: platform, observed property.
platform
A host or carrier for sensing equipment, such as a satellite.
Why it matters: The satellite identity is not the same as instrument or processing level.
Related: sensor, provenance.
Scale and support
spatial extent
The geographic domain covered by a dataset or study.
Why it matters: A scene boundary does not specify the support of each pixel value.
Related: spatial support, footprint.
spatial resolution
A stated measure of spatial detail or sampling spacing under a source convention.
Why it matters: Pixel spacing can differ from the effective sensor footprint or averaging support.
Related: spatial support, grain.
spatial support
The spatial domain or weighting over which a reported value applies.
Why it matters: A plot average and a satellite pixel are not equivalent because their centers overlap.
Related: footprint, aggregation.
footprint
The spatial area associated with an observation, sensor response or extracted window.
Why it matters: A bounding box may approximate the footprint without defining the full response function.
Related: spatial support, spatial extent.
CRS
A coordinate reference system defining how coordinates relate to Earth or another reference frame.
Why it matters: Native raster CRS differs from the CRS used to publish geographic footprints.
location
A position or geometry associated with a thing or observation.
Why it matters: A coordinate alone does not establish resolution, extent or support.
Related: CRS, spatial support.
temporal extent
The overall period covered by a dataset or study.
Why it matters: A year's coverage does not mean each value represents a yearly average.
Related: temporal support, interval.
temporal resolution
The time detail, spacing or cadence stated by a source.
Why it matters: Daily reporting does not prove daily averaging or complete daily coverage.
Related: temporal support, timestamp.
temporal support
The instant, interval or temporal weighting represented by a value.
Why it matters: An instantaneous acquisition and a daily mean differ even on the same date.
Related: instant, temporal extent.
timestamp
A recorded time marking an event or record under a stated convention.
Why it matters: One timestamp may indicate acquisition, publication or interval midpoint; inspect its role.
Related: temporal resolution, temporal support.
instant
A zero-duration temporal position under an explicit observation model.
Why it matters: Equal start/end acquisition times must not be turned into invented durations.
Related: interval, temporal support.
interval
A period with temporal boundaries.
Why it matters: An interval's boundaries do not automatically tell whether observations are averaged or sampled within it.
Related: instant, aggregation.
scale
A family of distinct properties describing organization, support, resolution, extent or analysis.
Why it matters: Scale is not one universal ladder and its meaning must be specified.
Related: biological scale, support.
biological scale
A study-specific description of biological organization or processes.
Why it matters: Molecular, anatomical, taxonomic and ecological organization intersect rather than forming one ranking.
Related: ecological organizational level, scale.
ecological organizational level
A category such as organism, population, community or ecosystem under an ecological model.
Why it matters: It is not automatically a spatial grain or taxonomic rank.
Related: population, spatial resolution.
grain
The size or detail of sampling or analysis units under an explicit usage.
Why it matters: Different disciplines use grain differently; document whether it means support, spacing or another property.
Related: spatial resolution, support.
extent
The overall domain considered in space, time or another specified dimension.
Why it matters: Always state the dimension rather than using extent as a synonym for resolution.
Related: spatial extent, temporal extent.
support
The domain, membership or weighting represented by an observation or statistic.
Why it matters: A support transformation is a scientific decision, not merely file-format conversion.
Related: spatial support, temporal support.
aggregation
Combining observations using a specified grouping, weighting and statistical rule.
Why it matters: A mean can change support and uncertainty; provenance must retain that change.
Related: upscaling, transformation.
upscaling
A stated transformation from one support or resolution to a coarser or broader one.
Why it matters: The statistical and scientific assumptions must be justified for the target use.
Related: aggregation, cross-scale transformation.
downscaling
A stated transformation to finer support or resolution, often requiring a model.
Why it matters: Finer output cells do not create new independent observations.
Related: upscaling, model judgment.
cross-scale transformation
A transformation connecting distinct organizational levels, supports or analysis scales.
Why it matters: A leaf-to-pixel relationship needs explicit scientific justification, not a chain of labels.
Related: support, derivation.
Quantitative semantics
variable
A specified characteristic represented in data, ideally with entity, property and context.
Why it matters: The same column name can represent different quantities and methods.
Related: trait, measurement.
quantity
A property that can be expressed by magnitude under a measurement system.
Why it matters: The quantity being measured is distinct from a numeric value and its chosen unit.
Related: quantity kind, unit.
quantity kind
A category of comparable quantities under a stated quantity system.
Why it matters: Equal dimensions alone do not make all quantities the same kind.
unit
An agreed reference used to express a quantity value.
Why it matters: Convertible units do not establish biological meaning or variable equivalence.
Related: dimensionless, numeric convertibility.
dimensionless
Having no physical dimension in the stated quantity system; often expressed with unit 1.
Why it matters: Different dimensionless quantities can have entirely different meanings.
Related: unit, quantity kind.
calibration
A documented relationship used to relate an instrument's encoded or indicated values to a target quantity.
Why it matters: Scale and offset must come from source evidence; applying them does not prove ecological validity.
Related: quality control, transformation.
quality control
Explicit checks and selection rules used to assess or filter data for a purpose.
Why it matters: A passed mask predicate is not universal proof of scientific suitability.
Related: calibration, uncertainty.
missing value
An explicit absence or unavailability of a value, including documented sentinels.
Why it matters: Zero can be a valid value or nodata depending on source evidence.
Related: quality control, needs evidence.
uncertainty
A documented characterization of limited knowledge about a result or decision.
Why it matters: Measurement uncertainty, model confidence and missing evidence are different things.
Related: confidence, measurement.
confidence
A stated assessment of certainty under an explicit interpretation.
Why it matters: An agent or curator score is not automatically a calibrated probability or measurement uncertainty.
Related: uncertainty, model judgment.
Provenance and harmonization
provenance
Information about the entities, activities and agents involved in producing a result.
Why it matters: Recorded history enables inspection but does not prove the activity was scientifically justified.
Related: lineage, derivation.
lineage
The chain of input identities and transformations leading to a product.
Why it matters: Streaming hashes preserve evidence identity without reconstructing discarded observations.
Related: provenance, transformation.
entity/activity/agent
PROV categories for things, production/use activities and responsible actors.
Why it matters: A scientific concept, its measurement and the software producing a result play different roles.
Related: provenance, instance.
derivation
A relationship recording that a product depends on another entity or result.
Why it matters: A recorded derivation does not validate the scientific assumptions used.
Related: transformation, provenance.
transformation
A specified operation producing a new result from inputs.
Why it matters: Scientific changes require versioned recipes and review; preprocessing has a separate identity.
Related: derivation, normalization.
harmonization
An evidenced process making data suitable for a stated joint interpretation while preserving source meaning and differences.
Why it matters: It can require conversion, explicit noncomparability or abstention; it is not automatic universal equivalence.
Related: standardization, scientific comparability.
standardization
Conformance to a specified naming, structure, unit or procedural standard.
Why it matters: Standardized fields can still encode different scientific meanings.
Related: harmonization, standard.
normalization
A context-dependent adjustment such as numerical scaling or structural normalization.
Why it matters: State whether it changes scientific meaning or is model preprocessing; the word alone is ambiguous.
Related: transformation, embedding.
semantic equivalence
An evidenced claim of the same relevant meaning under stated resources, versions and context.
Why it matters: Label similarity, shared units and spatial overlap are insufficient.
Related: equivalent, scientific comparability.
numeric convertibility
The existence of a supported mathematical conversion between numeric expressions, such as units.
Why it matters: A valid conversion does not establish that two variables measure the same thing.
Related: semantic equivalence, calibration.
scientific comparability
Suitability of observations for a specified scientific comparison under reviewed assumptions.
Why it matters: Meaning, methods, support, time, uncertainty and rights can all matter.
Related: semantic equivalence, support.
AI-ready data
Data with explicit structure, meanings, rights, provenance and interfaces suitable for a stated computational use.
Why it matters: A numeric vector alone does not guarantee scientific suitability or universal model understanding.
Related: representation, embedding.
representation
A machine-usable encoding of harmonized observations under an identified method.
Why it matters: Representation preprocessing is distinct from scientific harmonization.
Related: embedding, provenance.
embedding
A numeric representation generated by a specified encoder and preprocessing contract.
Why it matters: Shared vector dimensions do not imply a universal latent space or native LLM interpretation.
Related: representation, AI-ready data.
core harmonization
Reviewed transformations whose assumptions are explicitly shared across applicable tracks.
Why it matters: Promotion to the core requires scientific review; one track cannot impose its interpretation on others.
Related: track-specific harmonization, harmonization.
track-specific harmonization
Reviewed transformations scoped to a particular scientific question or track.
Why it matters: Shared sources do not make track-specific rules universally valid.
Related: core harmonization, scientific comparability.
evidence
An identified source, observation, deterministic derivation or explicitly classified judgment supporting a claim.
Why it matters: Evidence identity and scope matter more than a model's confident wording.
Related: source-documented evidence, model judgment.
source-documented evidence
A statement or annotation explicitly supplied by a source, retained with its location and resource identity.
Why it matters: An adapter's inferred description must not silently become a source statement.
Related: observed evidence, evidence.
observed evidence
Evidence actually encountered in data, with a declared observation or window scope.
Why it matters: An extracted window hash does not cover the entire remote asset.
Related: source-documented evidence, lineage.
deterministically derived evidence
Evidence produced by a specified reproducible rule from identified inputs.
Why it matters: A parser-derived fact remains distinct from a direct source annotation.
Related: evidence, transformation.
model judgment
An interpretation proposed by a model and recorded with model/prompt/output provenance.
Why it matters: It can be wrong and must not be promoted to source metadata or automatic approval.
Related: human decision, confidence.
human decision
A recorded person's explicit review outcome tied to exact scientific content.
Why it matters: A pending review is not an approval; local records do not authenticate identity by themselves.
Related: model judgment, provenance.
needs evidence
A state indicating that a required claim lacks adequate evidence.
Why it matters: Missing information is not automatically a demonstrated scientific failure.
Related: evidence, semantic gap.
semantic gap
A documented unmet meaning, relation or representation need in a defined comparison.
Why it matters: A local gap is a research question, not proof that no ontology anywhere addresses it.
Related: semantic composition, cross-scale transformation.
Benchmark interpretation
Ontology research framework · Dataset framework · Computed diagnostics
Explicit ontology use
Semantic infrastructure explicitly declared by a source: concept or taxon IDs, controlled vocabularies, unit systems, standard names or observation standards. An identifier is evidence to inspect, not proof of correct annotation or equivalence.
Ontology potential
The scientific meaning that could be mapped responsibly because definitions, methods, context and support are recoverable from evidence. A proposed assessment profile, not the current metadata-field presence score.
Semantic diversity
The range of distinct semantic dimensions involved in a dataset or question, such as taxa, traits, communities, environment, methods, sensors, space, time and support. It is different from record count, ontology count and semantic completeness.
Ontology leverage
A project research profile of semantic richness, recoverable meaning, connectivity, machine accessibility and evidence quality. It is not yet a calibrated scalar score. Low explicit ontology use can coexist with high potential and high leverage.
Breadth
The range of meanings a resource addresses. Current inventory breadth is a provisional 21-dimension assessment, not measured term coverage.
Depth
Specificity within a domain: hierarchical distinctions, relational structure or annotation detail. These require different tests; raw class count cannot replace them.
Composability
Whether selected concepts and relations can jointly express an intended meaning without incompatible commitments or loss. A shared label or namespace does not establish it.
Complementarity
Additional meaning supplied by another resource or dataset for a specific question. Assessed-scope gains are diagnostics; scientific complementarity requires evidence review.
Ontology accumulation
A proposed experiment measuring additional semantic coverage as resources are added. Early saturation and continued specialist contributions are alternatives to test, not results established by the current portfolio heuristic.