Free entity SEO learning

Describe entity attributes with entries and evidence

Entity attributes and predicates

An attribute states something about an entity. A predicate expresses the relationship in a statement. Describe the entry precisely, preserve its conditions and source, and distinguish a literal entry from another entity.

A desk lamp beside a workshop model, blank specification cards, a ruler and a notebook.

Start with a complete subject, relation and entry

Begin with the subject: which thing does the claim describe? Then choose the property or relation and its entry. A product can have a model name, a manufacturer and a measured width. The manufacturer is another organization; the width is a quantity whose unit affects meaning.

A predicate gives the relationship a direction. A product has a manufacturer; a manufacturer makes a product. Those statements concern related things but are not interchangeable property labels. Check the vocabulary’s documented meaning before selecting a property simply because its name looks familiar.

Identify what the entity is before attaching a claim. Two products with similar names may need separate records; a spelling match cannot settle identity.

A blue vase beside a ruler, finish samples and a reference binder showing the same object.

Distinguish an attribute from an identified related entity

A plain name can be sufficient for a simple visible explanation, but it should not silently combine distinct manufacturers or products. A related entity needs its own identity when other statements must refer to it consistently. Use a stable record and evidence rather than creating a new entity for every spelling variation.

For Study Lamp L2, “180 millimetres” is a quantity entry with a measurement basis. “Alder Lighting” refers to a separate fictional organization. If other claims describe that manufacturer, give it one consistent record rather than a new record for every spelling. Resolve similar names before combining those facts.

A blue desk lamp beside a workshop miniature, an unmarked ruler and a reference book.

Preserve the entry form and applicable conditions

Literal values include text, numbers and dates. Specify the unit and measurement basis for quantities. Preserve language, currency, date format and applicable version when they affect meaning. A maximum load rating under stated test conditions should not become an unconditional claim about every use.

Unknown

Meaning: The applicable source does not establish an entry.
Fictional example: The lamp’s mass has not been supplied; do not enter zero.

Zero

Meaning: An applicable source expressly reports the numeric quantity as zero.
Fictional example: A teaching inventory may record zero included spare bulbs; this says nothing about lamp mass.

Negative finding

Meaning: A defined check supports an absence within its scope.
Fictional example: An example package list states that no carrying case is included. That is a packaging finding, not a numeric measurement.

Keep different variants and conditions separate

A product can be offered in matte blue and matte white without one finish contradicting the other. State which variant a claim describes. A number measured across the base differs from a number measured across the shade. Keep both descriptions if both are supported; do not average them into a new specification.

In this invented exercise, revision A gives the base width as 180 millimetres. A later revision B gives the base width as 185 millimetres for a revised base. Keep the version beside each entry. Until the page’s exact model and revision are settled, the correct published answer may be qualified or held for confirmation.

Two lamps on separate trays, each with its own folder and finish sample.

Read the fictional Study Lamp L2 record

Imagine an unbranded teaching product called Study Lamp L2. Its invented manufacturer is Alder Lighting. The values below are fabricated for a modeling exercise and are not buying advice or specifications of a real product.

Store each entry with its evidence source, version and applicable conditions. If a specification sheet changes, review the affected claims rather than changing every statement about the product indiscriminately. A missing entry remains unknown; it does not become zero.

Study Lamp L2 has manufacturer

Fact: Alder Lighting
Condition and evidence record: Invented organization in this exercise

Study Lamp L2 has width

Fact: 180 millimetres
Condition and evidence record: Base width; fictional specification revision A

Study Lamp L2 has power rating

Fact: 8 watts
Condition and evidence record: Rated input; fictional specification revision A

Study Lamp L2 has finish

Fact: Matte blue
Condition and evidence record: Example variant only

Attach provenance to each claim

A source record should identify the document or responsible owner, retrieval or confirmation date, version and exact scope of the supported statement. Distinguish a manufacturer’s specification from an independently measured result. Both may be useful, but they establish different kinds of evidence.

For a measured number, record what was measured and how. For a manufacturer statement, identify the manufacturer document and version. Both can be retained without pretending they are the same kind of observation. See claims, numbers and provenance for the full evidence record.

An open reference binder, a document folder, a clock and a magnifying glass.

Resolve conflict before publishing a stronger statement

When sources conflict, retain both observations and investigate the version, product variant, measurement method or date. Do not average incompatible values or choose the more attractive one. The published answer may need a qualification or may need to wait for confirmation.

A review can conclude that two observations concern different variants, that one supersedes the other, or that the disagreement remains unresolved. Record the reason. If the source cannot settle the intended product revision, withhold the unconditional claim and use the entity evidence conflict method.

Choose a vocabulary from meaning and expected entries

Schema.org’s data model distinguishes types and properties and allows property values with documented expected forms. W3C’s RDF model distinguishes subjects, predicates and objects. Neither source establishes the truth of a particular product specification. Check meaning and evidence before validating syntax.

Choose a property after the visible statement is precise. Check the documented meaning and expected entry form for that property. A property named similarly to your editorial label may describe a different relationship. The property evidence matrix connects the selection to supported facts; graphs, taxonomies and ontologies explains the wider representation.

Maintain a claim record as the page changes

This static record shows the effect of the invented revision B. The source labels, dates and owner below belong to the exercise, with no real product specification implied.

A blue sheet clipped into an open binder beside four separate blank record cards.

Claim requiring review

Subject ID: L2-A, the fictional revision A lamp.
Predicate meaning: Base width measured across the base.
Entry or object ID: 180 millimetres in revision A; 185 millimetres in revision B for the revised base.
Unit: Millimetres.
Condition: Use the entry for the matching product revision only.
Source: Invented specification sheet L2-SPEC.
Source version: A retained as history; B supplies the new base description.
Evidence state: Fictional teaching entries; real product facts unknown.
Owner: Example product editor.
Review trigger: A changed base specification or confirmation that the page covers revision B.

Impact on the other claims

Review: Check manufacturer, rated input and finish against revision B.
Change decision: The exercise changes only base width. Keep the other entries unless the applicable source also changes them.
History: Retain revision A and the reason for selecting revision B; do not silently erase earlier evidence.

Use evidence and expiry checks when a source changes. The stable entity ID method keeps references traceable.

Review the published explanation in context

Check that every record entry and sentence still refers to the intended product and variant. Put conditions close to the entry they govern. Keep units visible in tables, and do not rely on a distant heading to supply a necessary limit.

Check entity identity across pages when the same product appears elsewhere. A consistent record still needs each page’s relevant qualifications.

Sources and next steps

Read the semantic SEO learning hub and the entity SEO overview for the surrounding topics.

Official references: Schema.org data model, RDF concepts. Examples on this page are illustrative, not measured customer results.

Magnifying glass over grouped source pages and related subject markers.