Datasets

The datasets are stored as database views, which are virtual tables that are based on an underlying SQL definition. Dataset retrieval is often the biggest contributor to the completion time for a detection process. Some datasets use group by expressions to perform aggregate analysis on a set of records. Some datasets are also created by performing union operations on multiple queries.

Contributors to dataset retrieval time include the following:

• Volume of data being searched in the dataset.
• Extent to which conditions can be applied to filter the returned records.
• Number of PDM tables that are joined together in the dataset.
• Execution plan that the database uses to run the dataset query.