A collection of numbers or words, generally centered around a single particular topic or subject.
Datasets are most commonly compiled and stored in a tabular format (for example, as a CSV file).
Typically, each row corresponds to one entity, and each column to a property. The intersection of each column and row contains the relevant property’s value.
Datasets are typically compiled from observations and measurements made by people, recorded through sensors, or otherwise tracked by software systems.
Datasets can also be generated for the purpose of training a machine learning model, or providing a dataset that approximates a real-world one, without exposing potentially sensitive real underlying data.