Metadata Driven Data Governance

From MIKE2 Methodology

Taking a Metadata-Driven approach to Data Governance can greatly streamline a very complex process.

Contents 1 Overview 2 Metadata Artefacts for Data Quality Tracking 3 Data Governance Practices Defined as Metadata 4 Metadata Reuse Policies

Overview

Metadata Management is a critical aspect of any Data Governance initiative. In addition to establishing Data Governance policies and procedures, organisations must have methods, quality assurance checks, and initiatives to ensure that all aspects of metadata management are performed. Some of the key benefits of this approach are shown below.

BUSINESS/TECHNICAL VALUE	BENEFIT MEASURES
Improved business decision making	Data quality is improved, which provides business users with more accurate systems and reports
Reduction of technical-related problems	Improved data quality reduces many system-related problems and technical expenses
Increased system value to the business	Business users are likely to make better decisions if they are aware of possible errors skewing report numbers
Improved system performance	As data quality improves, system errors are reduced, which improves system performance

Metadata Artefacts for Data Quality Tracking

Metadata Management is a critical component to the enterprise data quality tracking:

• Having access to the metadata for all steps in the life cycle allows the data quality capability to highlight points of failure, origination points and redundant data.

• Data quality metrics should be stored in the metadata repository and retained throughout the life of the monitored system. This allows companies to monitor their data quality over time and to determine whether the quality is improving or declining.

• Data Re-Engineering artifacts are ideally stored in a metadata repository:
• Data Standardisation– metadata mappings from source to standardised model
• Data Correction – updates for future re-engineering processes
• Data Matching and Consolidation – metadata of source to target mapping rules for data matching/consolidation
• Data Enrichment – new enhancement to the meta-model to reflect any requirements for data enrichments

Having quality data is a key to accurate regulatory compliance reports and only through managing metadata can organisations achieve and maintain acceptable levels of data quality. Data quality tracking is valuable to many people within an organization, including corporate executives, project managers, database administrators, programmers, data modellers, business analysts, and business users across the enterprise.

Data Governance Practices Defined as Metadata

When data quality issues are discovered, the metadata will provide the information for the point of contact to begin the data correction and from there data governance plays a significant role. Proper Data Governance practices will enable information management group to answering questions such as, "Is this process or data element used in one LOB the same as that in another business unit?”; if so, "What are the policies and rules regarding sharing them?" Or "Have business unit owners who can authorize changes to shared metadata been identified?"

Proper definition of these practices means that is modelled in a metadata-driven fashion. It starts with the standardisation of business vocabularies and data dictionaries across data sources. It facilitates the effective use of data, quality of the data, improved data accessibility as well as oversight on business conditions derived from data analysis.

Organisational Responsibilities for Information Development are also important. Data stewardship rules, for example, enable the most relevant abstraction to be presented to various users—from developers to IT management, from auditors to executives. By increasing the level of stewardship in a data governance practice, organisations gain significant insight and control into the states of business and take timely, decisive actions to exploit new opportunities or mitigate potential risks.

Metadata Reuse Policies

In terms of governance area, a key issue for metadata management is the establishment of a set of metadata reuse policies. The reuse assumes a level of governance in which owners are identified and the reusers of the services and components agree to accept new versions based on collaborative processes between the owners and users.

The issue of governance and the ability and willingness to share in common definitions and new versions are at the heart of the ROI from metadata management initiatives. The greater the commonality and sharing across the users, the greater the ROI on the metadata management efforts, especially if these replace redundancies. Reuse initiatives of Services Oriented Architectures have at their core metadata sharing and reuse.

The value of metadata reuse is significant and the formula for calculating its value is simple - whatever effort, cost or time it took to initially define and build the metadata artefact, subsequent reuse saves that effort, cost or time, minus the additional overhead of maintaining the shared artefact across multiple users. Maintenance costs are also generally reduced when reuse programs are instituted, because there are fewer artefacts to maintain, and quality and consistency improve, as does responsiveness.

IT productivity improvements translate into reuse of data models and components. Reuse is a bootstrap operation requiring organisational dynamics, business requirements analysis and automation. Metadata-driven design incorporates aspects of all three.