Attribute–Value System¶
An attribute–value system describes each object through a shared set of typed attributes, assigning a value to every object–attribute pair in its complete form.
Core Idea¶
An attribute–value system describes objects \(U\) by a shared set of attributes \(A\). Each attribute \(a\) has a domain \(V_a\), and a complete valuation assigns an admissible \(f(u,a)\) to every object–attribute pair. Objects form rows, attributes columns. A decision/class column is an optional specialization; equality of values on a chosen subset of attributes yields a rough-set indiscernibility relation.[^ref-275b7771b289]
Scope of Application¶
Attributes may be binary, multivalued or numeric; “more than two values” is permitted, not required. An EAV triple layout can store the same facts but is a storage encoding, while an ER diagram supplies schema types without object-wise values. The live action-prescribing Decision Table is also not the rough-set information table. Missing or inapplicable entries require explicit semantics instead of being silently treated as known values.[^ref-c1005f872167]
Clarity¶
Name the unit of observation, common columns, each domain, and whether every cell is known. Then identify any target column and the attribute subset used for comparison. Two cases equal on one column may differ on another; a class label does not by itself make condition values deterministic.
Manages Complexity¶
UCI Car Evaluation records 1,728 cases with six categorical inputs—buying, maintenance, doors, persons, luggage boot and safety—and an optional class target. Buying has four listed levels and safety three; no values are missing. Mapped back: cases are objects, six common typed properties are attributes, and class is a distinguished analytical role rather than core identity.[^ref-010fa2178e0e]
Abstract Reasoning¶
UCI Poker Hand instead records 1,025,010 five-card hands using five position-indexed suit/rank pairs, ten predictive attributes, plus class. Suit values 1–4 and ranks 1–13 fill different domains. Mapped back: a hand is the object; the repeated card positions are shared attributes; order matters in this dataset. Unlike car properties, the schema repeats the same two attribute types across five positions.[^ref-40db51084a80]
Knowledge Transfer¶
Fixed columns make cases comparable but can flatten heterogeneous or missing detail; richer unknown-value handling preserves distinctions while complicating equality and indiscernibility.[^ref-c1005f872167] The typed object–attribute valuation remains domain-specific. Live Representation is its reviewed strict genus for the complete-valued base identity, not a claim that missing cells are ordinary values.
[^ref-275b7771b289]: Andrzej Skowron et al., “Rough Sets: Past, Present, and Future” (2018), formal information-system and decision-system definitions. https://pmc.ncbi.nlm.nih.gov/articles/PMC6244804/ [^ref-010fa2178e0e]: Marko Bohanec, “Car Evaluation,” UCI Machine Learning Repository. https://archive.ics.uci.edu/dataset/19/car+evaluation [^ref-40db51084a80]: Robert Cattral and Franz Oppacher, “Poker Hand,” UCI Machine Learning Repository. https://archive.ics.uci.edu/dataset/158/poker+hand [^ref-c1005f872167]: Marzena Kryszkiewicz, “Rough Set Approach to Incomplete Information Systems,” Information Sciences 112 (1998), abstract; full text inaccessible in author check. https://www.sciencedirect.com/science/article/pii/S0020025598100191/pdf
Relationships to Other Abstractions¶
Current abstraction Attribute–Value System Domain-specific
Parents (1) — more general patterns this builds on
-
Attribute–Value System is a kind of Representation Prime
An attribute–value system represents objects through a shared typed attribute-to-value mapping.
Hierarchy path (1) — routes to 1 parentless root
- Attribute–Value System → Representation → Abstraction
Neighborhood in Abstraction Space¶
Attribute–Value System sits in a sparse region of the domain-specific corpus (80th percentile for distinctiveness): few abstractions share its structure, so a faithful description tends to retrieve it precisely.
Family — Property Ontology & Code Smells (18 abstractions)
Nearest neighbors
- Data Model — 0.83
- Object–relational model — 0.82
- Relational Model — 0.82
- Moduli Space — 0.82
- Valuation (logic) — 0.82
Computed from structural-signature embeddings · 2026-10-08