SOFTWARE - DATA STANDARDS

Introduction

The sheep industry now has standards for its animal, wool and meat data terminology, as well as for storage and transmission of this data.

The development and implementation of these standards will aid in the faster uptake of data management software systems by sheep producers.

So far, development of the standards has focused on the software developers and service providers. A data dictionary, data architecture and a reference database have been produced for uptake by software developers.

The next stage of the standards project is to form an industry group to manage the standards and to work with service providers to develop messaging protocols for data transfer and to foster the uptake of the standards by industry.

Sheep producers are welcome to view and explore the standards, but be aware that the main site has been developed so far with software developers in mind.

For further information contact Tim Dyall 02 6776 1463

For Industry

A glossary of data terms and their definitions will be published in the near future.

It is hoped that Industry groups such as MLA and SGA will participate in the development of post-farm and interfacing protocols.

For wool producers, there is enormous opportunity for inclusion of electronic classer specifications, e-bale identification and tracking and retrieval of wool sale information from brokers.

For meat producers, feedback sheets from abattoirs can easily be incorporated.

For all producers, inclusion of animal health measures in the data standards will enhance Quality Assurance along the marketing chain across the entire industry.

We have identified areas where the transmission of messages (known as messaging) between producers and service providers is required – areas such as classing, testing, genetics, carcase data and animal health. If further funding can be sourced from industry, XML messaging standards for common industry documentation such as electronic wool classer specification, vendor declarations (chemical and dark fibre risk), and wool test data from small wool test houses can be developed. This will aid in the adoption of the standards by industry.

Development of record capture, management and analysis software is complex, as is evident from the large number of standardised terms now available in our database.

There are already commercial software programs developed for a variety of sheep industry applications. The Sheep CRC recommends that producers choose one or more of these to suit their end use.

Software users should ask their software providers and providers of data (e.g. fibre measurements) whether they conform to the new data standards. While current software and data is unlikely to conform to the standards right now, as new versions are developed, the standards should be incorporated to ensure data is compatible and easily transferred.

For Software Developers

The Data Standards Project has produced a set of standards that will provide the sheep industry with a common language for on-farm data storage and messaging.

The draft standards were developed using the following "rules" and philosophies:

The scope of the project was limited to the collection and storage of data on sheep farms and was limited to fields that are commonly used. Some areas that did not make it into this first draft include prime lamb production and animal health diagnostic services. Information that is generated post farm (for example wool sales invoices and appraisals, abattoir feedback sheets) and is viewed and used as historical information for farm planning and management was not included.
Field names are kept as descriptive as possible, with the first letter of each word in capitals and a 3-letter abbreviation used at the start of each field name.
All look-up tables are prefixed with LK.
Codes for look-up tables are made descriptive, using capital letters and a maximum of 12 characters, where possible. Number codes have been avoided unless widely adopted by industry.
The data architecture contains more than 30 main tables and nearly 50 look-up tables. Look-up tables are used extensively so that storage of information is standardised as much as possible. Very few of the look-up tables are user defined.
Measurement data, where the type of measurement is variable, is stored in a vertical format with one row per data item. For example, the EBV record is vertical and each row contains a single EBV value. New EBV types are easily be accommodated in the database by adding an EBV definition to the EBV's look-up table.
Data which has a fairly fixed format is stored in a horizontal format, where each row holds multiple data items. For example, the Animal record is horizontal and contains numerous pieces of information about the individual animal in a single row.

Data Standards

Download	Resource
Schema (Visio)	The Microsoft Visio document defines the data standard as an Entity-Relationship (ER) diagram. The diagram includes look-up table values in the table notes fields. This document is the source of the reference databases and the look-up table data files that can be imported into existing applications.
Schema (DDL)	The Data-Definition-Language (DDL) file contains the schema for the full reference database in a SQL-92 dialect. Import this file into a SQL database to generate the tables and constraints for the database.
Look-up data in comma-delimited text files	The look-up table files contain standard codes for traits, standards for measurement units, and many other entities that can be recorded in the databases.

Reference databases in Microsoft Access 2000

The full database contains tables to support data management for a broad range of animal information applications. To make it easier to use the database design, individual databases have been generated that contain subsets of tables which apply to certain applications:

Application	Definition
Full Database	Contains all tables for all applications.
Animal Management	All tables with descriptions, measurements, attributes, treatments and events for the management and sale of an animal or product from the animal. It includes all tables with information about the paddocks and the property as well as mob information required for vendor declarations, quality assurance and traceback.
Genetics	Tables with information required for genetic evaluation of animals. It includes all mating type tables and tables with phenotypic measurements.
Wool	Tables with information required for the wool classer specification sheet and for measurements and scores recorded when fleeces are classed.

Further information on the data standards project: email Tim Dyall CSIRO LIvestock Industries or call 02 6776 1463

Sheep Industry Data Standards database implementation developed by: Hutchinson Software