connection
Let an agent handle connections by type, parties, direction and evidence, and avoid asserting connections without support.
Bundle → Layer → Finding → Questions Filled
4 bundles · 8 layers · 8 findings · 16 questions
Type What kind.
Kind
Causal, similarity, social.
Kind
Connection type.
- What kind of connection is it? definition
- Between which entities? definition
Direction
One-way or mutual.
Direction
Direction.
- Is the connection directed? definition
- Is it symmetric? definition
Evidence How known.
Source
Evidence.
Evidence
Evidence for the connection.
- What evidence supports this connection? provenance
- How strong is it? measurement
Alternatives
Other explanations.
Alternatives
Alternative explanations.
- Could the apparent connection be coincidence? boundary
- What else could explain it? boundary
Networks Many connections.
Mapping
Graphs.
Mapping
Network mapping.
- How are connections mapped? definition
- What does the network show? definition
Change
Over time.
Change
Changes.
- How have connections changed? provenance
- When? provenance
Privacy Personal relations.
Data
Relationship data.
Data
Data handling.
- Is relationship data handled lawfully? boundary
- Who can see it? boundary
Fairness
Association.
Fairness
Avoiding guilt by association.
- Is anyone judged unfairly by association? boundary
- How should that be avoided? action
Classifiers Filled
- Family
- Thing Registry
- Category
- Cross-cutting context
- Entry kind
- noun.relation
- Plane
- XCT
- Domain
- XCT.REL
- Facets: substance
- abstract
- Facets: origin
- conceptual
- Facets: agency
- inert
- Facets: mobility
- not-applicable
- Facets: scale
- not-applicable
- Facets: affordances
- observable
- Other names and narrower kinds
- connexion, connectedness, alliance, bond, silver cord, linkage, communication, concatenation, involvement, implication, inclusion, comprehension
What it is Filled
A relation between things or events, as when one causes the other or they share features, or a link, bond or alliance between people or entities.
Why it exists Filled
Let an agent handle connections by type, parties, direction and evidence, and avoid asserting connections without support.
Distinguishing features Filled
- Relation between two or more things
- Can be causal, similarity-based or social
- Needs evidence
- Direction may matter
What robots and AI may and may not do Filled
Must not
- Assert a causal connection without evidence.
- Infer connections between people, such as associations or relationships, and publish them without evidence or consent.
- Use networks of connections to profile or target people unlawfully.
- Invent links between events to support a conspiracy theory.
Only with a human decision
- Publishing an inferred connection between named people.
May
- Record and describe connections between things with their evidence.
- Distinguish a correlation from a causal connection in reports.
Moral aspects Filled
- Guilt by association harms people who are linked to others without cause.
- Confusing correlation with cause leads to bad decisions in health and policy.
Who is affected
- People linked in data
- Readers who rely on the claims
Owners Filled
Steward
Nobody: a general concept; the keeper of a dataset answers for the connections recorded in it.
Links to other meta-models Filled
parent
- oewn-00032220-n - registry parent class
related
- relation - abstraction
- causality and kinship - types
- network - structure
- correlation - weak link
What else AI and robots need to interact with it Filled
Identity and identifiers required Filled
- Vercy registry: vr.tr.connection
- Open English WordNet: oewn-13813601-n
Direct properties not applicable Not applicable
Not applicable
Plane XCT: no invented physical properties.
Recognition optional Filled
- Links in diagrams and graphs
- Stated relationships
- Physical connections are a related but different sense
- Not a thing; seen in networks, links and relationships.
Capabilities and actions required Filled
- identify and describe it
- evidence it
- map networks of connections
- break or form connections
Hazards and failure modes required Filled
- Claiming connections without evidence
- Guilt by association
- Revealing private relationships
Standards and interfaces required Filled
- Data protection for relationship data
Context of use required Filled
- Linking entities in reasoning, networks and relationships.
- causal connections
- similarity connections
- social bonds and alliances
- physical and network links
Sources Filled
- Open English WordNet 2024, synset oewn-13813601-n - sense and gloss of the concept
- Wiktionary: connection - word senses and usage
Open questions
- How should connection types be modelled?
- How should evidence strength be attached?
- How should social connection data be protected?
Machine files
Provenance
thing registry research (pass 2) · unreviewed
Built from: models/things/publications/thing-connection/spec.json