Percentages: closing the open note on ont00000845

[UtopianKnives]

INTRODUCTION

CCO 2.1 (versionIRI .../2026-04-04/...) has no way to express a
percentage as a value. There is no percent or proportion measurement unit,
and no percentage-value data property.
The rdfs:comment on ont00000845 Proportional Ratio Measurement
Information Content Entity already raises exactly this, weighing a
Percentage subclass against a percentage-value data property. So this is a
note you have already written; I am bringing a use case for closing it.

What I checked, so you know what the claim rests on

All 30 transitive subclasses of ont00000120 Measurement Unit, and all 269
named individuals typed to any of them, matching rdfs:label,
skos:altLabel and the SI unit symbol annotation against percent, per cent,
%, dimensionless, ratio, proportion, ppm, parts per, fraction, basis point
and per mille.
One match, and it is a false positive: "Per Cent" inside Kilogram Force Per
Centimeter Square Measurement Unit. Of the 269 unit individuals, 28 carry
an SI unit symbol and % is not among them.

The use case

CUAD is a corpus of 510 commercial contracts with clause-level annotation by
lawyers, released CC BY 4.0. Measuring the annotated span text:
22 of 40 clause types carry a duration, monetary amount or percentage in
at least 25% of their spans, and 16 of 40 in at least half;
percentages concentrate — 91 of the 166 annotated Revenue/Profit Sharing
spans carry one, and they recur across Audit Rights, Minimum Commitment,
Change Of Control and Liquidated Damages.
Those counts come from regular expressions, so they are a floor: amounts
written out in words are missed.
The point of the comparison is that durations and currency amounts from the
same corpus are already expressible, using the existing time and currency
unit individuals. A percentage is the one common contract value with no
licensed form.

Possible resolutions, ranked as I see them

A has percentage value data property with domain ont00000253,
matching the shape of the existing twelve — the second option your own
ont00000845 comment contemplates.
A percent unit individual alongside the existing 269, typed to a
dimensionless or ratio unit class.
A documented convention: the proportion is carried by
has decimal value on the bearer of a Proportional Ratio Measurement
ICE, normalised to the proportion (0.05, not 5), with no unit asserted
on the grounds that a ratio is dimensionless. Nothing minted — but the
normalisation needs stating somewhere, because 5 and 0.05 are both
defensible readings of "5%" and a consumer cannot tell them apart from
the literal.
Option 3 may already be the intent, in which case saying so in the
ont00000845 definition or a design pattern would settle it without adding
a term.

One question about 3.0

The milestone lists the incorporation of QUDT. Is a percent or
proportion unit expected to arrive with it? If so this is already handled
and the only thing worth doing is recording the interim convention.
I am happy to contribute a worked serialization to the design-patterns
directory for whichever route you take.