Data and evidence

A primary source is a record produced by a person or body with direct knowledge of the fact it records, at or near the time, as opposed to a later account that reports, interprets or repeats it.

Beneficial ownership, in anti-money-laundering law, is ownership or control of a legal entity or arrangement traced to the natural persons at the end of the chain; each regime sets its own ownership threshold, commonly 25%, and control test.

Data provenance is the record of where a piece of data came from and how it was produced: the source artefacts, the activities that generated or changed it, the people or systems responsible, and when each step happened.

Evidence dates are the distinct dates attached to a fact and the evidence for it: when the fact took effect, the date a source states it as of, when the source was filed or published, captured, recorded and last verified.

Temporal validity is the period during which a fact is true in the world, its valid time, recorded as a start and an end so that the fact can be answered for any date rather than only for today.

Point-in-time data is data that can be reproduced exactly as it was known on a past date, including values later corrected and entities that later disappeared, so that analysis of that date uses only information available then.

Bitemporal data is a record of facts kept on two independent time axes: valid time, when each fact was true in the world, and transaction time, when the database held that version of it.

Data freshness is a measure of how recently a value was last confirmed against its source, relative to when it is used, usually expressed as the age of the evidence and judged against how quickly that kind of value changes.

Source reliability is an assessment of how dependable a source has proven to be, based on its record, access and conduct, graded separately from the credibility of any single item of information it supplies.

Triangulation is the practice of checking a claim against two or more sources, methods or kinds of evidence that are independent of one another, so that their agreement cannot be explained by a shared origin or a shared error.

A confidence score is a stated measure, numeric or ordinal, of how strongly the available evidence supports a claim, a match or a judgment; its meaning is set by the method that produces it.

Entity resolution is the process of deciding which records, within or across datasets, refer to the same real-world entity (a company, fund, institution or person) and linking or merging them while keeping distinct entities apart.

Precision and recall measure the quality of a classification or matching process: precision is the share of items it flagged that are correct, and recall is the share of all correct items that it flagged.

Data quality dimensions are the measurable characteristics used to judge whether data is fit for a given use; a commonly used set is accuracy, completeness, consistency, timeliness, validity and uniqueness.