Overview

Attributes are individual data elements such as an email address, user agent, IP address, postal code etc. Darwinium uses Attributes as a means to reference data elements with an inferred meaning or purpose.

Attributes are organised into three categories.

Identifiers

Identifiers are attributes whose values are unique enough to identify a user or a piece of data with a reasonably high degree of certainty.

For example, Device ID, Email Address, Telephone Number, (Street) Address and IP Address are all Identifier Attributes.

PDQ and Text DSH are also identifiers as they can uniquely identify images and blocks of text.

Because of their uniqueness, Identifiers act like "Indexes","Primary Keys" or "Pivots" on historic data when used within Features.Features are numeric values computed on event history over time. Many Features require one or more Identifiers to be selected to form an index for the Feature search query.

Subjects

Subjects are attributes that lack uniqueness but are extremely valuable to analyse over time (in past events).

Attributes

Attributes are generalised pieces of data that are useful within the current event, but not so useful when analysed over time (in past events). Attributes can be utilised within Decisioning but are not available for use in Features.

Data Types

Attributes (including Identifiers and Subjects) also have different data types similar to database columns, spreadsheet cells and variables in programming languages. Data Type is a classification that specifies what kind of data an attribute can hold and what type of relational, logical, mathematic or text operations can be applied without causing errors.

Integer

Integer Attributes are whole numbers (not fractional) that can be positive, negative or 0. Integer attributes are signed 64 bit integers and can represent whole numbers in the range [-9,223,372,036,854,775,808 to 9,223,036,854,755,807].

Real

Real Attributes can represent both fractional and whole numbers that can be positive, negative or 0. Real attributes are IEEE 64bit double precision floating point format.

Boolean

Boolean Attributes (often called "Flags") expresses a truth value which can be true or false.

String

String Attributes are an ordered series of unicode UTF-16 characters that comprise a body of text data. An example of a String Attribute is "Hello World" or "ハロー・ワールド". All string comparisons are case sensitive.

String Array

String Array Attributes are an ordered series of Strings. Strings can be accessed via their position in the series.

String Map

String Map Attributes are associative arrays (also known as key-value pairs). Each entry consists of a key which is the "name" of the String being referenced and a value that is associated with that key.

Data Groups

The Darwinium schema is hiearchical and groups datapoints in the schema.

The highest level categories of those groupings are explained below.