Data source refererce to the source from which a data is coming from, examples are: American Bureau of Statistics, Yahoo Finance, Wharton Research Data Services (WRDS), Thomson Reuters, etc.
Dataset refers to the data file provided by a data source which contain the data for one or multiple measures.
Dataset structure refers to the meta data for a dataset such as what measures are captured, unit of measur.
Link Type describes the type of link that exists between two or more variables.
A link describes a connection between two variables, the nature of this connection and how this connection was established. When the link is directional, it is read to the direction of LinkType. e.g. - firstVariable causes secondVariable, for Causal LinkType.
Measures are the actual values for variables and are expected to have following features:
· A unique identifier for each measure
· Dataset structure for the measure
Models refer to underlying technique(s) which establish the link between two variables. The model is required to have following specifications:
· Class of model such as statistical or mathematical model
· Type of model within a particular class
· Additional features within each model type
Variables exist as metrics to quantify concepts. Variables are required to have following features:
· A unique identifier for each variable
· The origin of the variables (references, people, etc.)
· Concepts associated with a variable, and link to those concepts
Causal: a variable causes another variable
Two variables are correlated to each other.
·Investigative Causal: there is a hypothesis that one variable cause another, the hypothesis needs to be tested.
NonCausal: Unable to establish a causal link between two variables.
One variable can be used as a proxy to the second variable.