Teradata Us, Inc.

Scott Shelton

Michael Chaves

Henry Fu

Jason Keith Figge

Leslie Mannion

Robert Baril

David Carmer

Harold R Lee

David Brophy

Anish Shah

Kasturi Belur

Claud Teeter

Christian Seifert

Dan Willoughby

Steven Kennedy

The subject mater herein relates to database management and, more particularly, to non-equijoin metadata. Various embodiments described herein provide systems, methods, software, and data structures to facilitate non-equijoins of data. Some such embodiments include a data store to hold metadata describing relationships between database tables. The relationships between database tables may include one or more non-equijoins of tables in the form of one or more table-column pairs of a first table being non-equal to one or more table-column pairs of a second table. In some embodiments, the metadata may be used by a query generator to build query statements including one or more non-equijoin.

Non-equijoin metadata

Tools for defining and using custom analysis modules are presented. A generic graphical user interface (GUI) custom analysis tool is presented to a user. The generic GUI custom analysis tool is adapted to receive selections and inputs from the user that define many-to-many relationships in a data store and that include user-defined measures. The inputs are used to generate an instance of a custom analysis module that is capable of being processed against the data store. The instance of the custom analysis module may be stored, recalled, executed, and/or modified from the generic GUI custom analysis tool.

Tools for defining and using custom analysis modules

Techniques are presented for generic data extraction. Metadata defines a data extraction element with reference to an identifier, a data source, and one or more rules. The metadata may be processed to populate the data extraction element when a reference is made within a data template to the identifier. The identifier may be used to import data to a template or to export data to a different template or service.

Techniques for generic data extraction

Techniques for applying segment conditions to measure results are presented. Row and column segment data is acquired from a database according to row and column definitions. A measure calculation is applied against the intersecting row and column segment data. Next, one or more combinations or conditions derived from the row and column definitions are used as a filter to produce filtered results.

Applying segment conditions to measure results

Techniques are presented for reusing business measures within a data store. A measure is associated with a measure type. Each measure is associated with a particular data store schema or schema element and includes metadata for fields and calculations against those fields. Requestors can select existing measures, create new measures, delete existing measures, and/or modify existing measures. A measure, when processed against the data store, produces results from a data store that are associated with the data store schema or the schema element and which conform to the fields and calculations defined within the measure's metadata.

Reusable measures

Growing community of inventors

Apex, NC, United States of America

Scott Shelton