Introduction
Literally “data about data”, metadata is a concept that is receiving increased visibility throughout the information systems communities, making the need for understanding metadata and its varied
roles across the design and use of systems and the information they capture. To make the data useful, to show where the data originated and is stored currently, to describe a host of descriptive
and technical attributes of that data – all can be considered the purpose of metadata. Management of this information becomes critical, and should be planned and implemented according to a strategy
that offers continuing value to the organization.
Although metadata is “data about data”, it really can apply to any information object (text, photos, audio/video, graphical, etc.) All information objects have three features in common: content,
context, structure, as explained in research standards developed for the Getty Library (http://www.getty.edu). Content relates to what the object contains or is
about; context is the “who, what, why, where and how” aspects associated with the object’s creation; structure is concerned with the set of associations among objects. All of these features can
be described as metadata, and may be of importance to one or more groups of users for that object.
Since metadata is diverse, and is used in many ways by a wide range of user types, the management of metadata is of extreme importance in any organization. Knowledge of some metadata management
best practices can lead to developing a robust and flexible approach to metadata management and to providing increased value to the information objects that are important to the organization’s
decisions.
Importance of Metadata
There are many forces at work in the drive toward better understanding and management of the various types of metadata. Some of these forces include:
- Number of legal and political directives concerning management and storage and access to personal data in both private and public sector databases (Sarbanes-Oxley, HIPPA, Freedom of Information
acts, etc.) - Improvement of access to services and data by the public (“e-business”, “e-government”)
- Increased recognition that organizations must make better use of their data for decision making and operations
- Ownership of data, understanding and managing data collection and storage and their use for specific purposes
Types and Functions of Metadata
There are several categories of metadata, each with its own definition and usage examples and user communities.
Metadata is not created on its own, often it is a direct by-product of the original source information systems that create, maintain and disseminate data. Also, metadata is created later and it can
be derived from system artifacts and the organization’s use of a system for operations or decision making. Without accessible and understandable metadata, information content may be meaningless to
the user.
Principles / Best Practices of Metadata Management
As in any other discipline, there are a host of best practices for metadata management that can assist an organization to establish and maintain a robust, enterprise view of metadata. Following are
some suggestions for managing metadata properly:
Develop a Metadata Administration Team – If an organization does not have an established data administration group, forming one should be the first goal. This is the first
and most significant step for any organization that wants to implement systematic metadata management. The metadata administration team would be part of any efforts concerning metadata in the
organization, from participating in the creation of a metadata strategy, adopting and refining metadata standards, choosing metadata management tools and implementing effective enterprise-wide
metadata management and understanding.
Develop a Data / Metadata Stewardship program – The leaders of Metadata Administration team should include the organization’s data / metadata stewards. Development of a
stewardship program is one of the core best practices in metadata management and can be the single most important indicator of a successful metadata strategy.
Create a Metadata Strategy / Policy – A good metadata strategy can assist an organization by raising the visibility of metadata as a companion resource to data /
information objects, and can become part of the organization’s enterprise architecture. The strategy’s most important factor is its ability to explain the value and context of the information
stored in the organization’s information systems, from both business and technical perspectives. To create this strategy, ask the following questions:
- What kind of content and “information” (meaning metadata) is needed to solve business problems / business questions / business issues?
- What technical issues / problems / questions can be solved by metadata?
- How can you get that content / metadata? How much will it cost to obtain / create that metadata?
- What types of metadata are needed for a particular user group?
- How will metadata be maintained, and updated?
- What standards should you follow? Should you adopt and modify an industry standard (Dublin Core, ISO 19115, PRISM, etc.) or should the organization create its own standards?
Adopt / Create / Refine Metadata Standards – Over the past several years various bodies have produced standards for metadata. These standards have much in common, but vary
in the degree of complexity and level of detail required to complete an instance of metadata. Each organization should evaluate the variety of “standards” that exist and use the most appropriate
of each for an organization-oriented standard that will be useful and maintained. Maintaining these standards is an essential best practice since much value can be stored in metadata, and losing
that value can create pockets of useless data within the organization.
Identify Appropriate Metadata Tools – There are many products that create and manage metadata, and each has its own purpose, platform, strengths and weaknesses. Choosing
the right tool set and implementing the tools appropriately should be the responsibility of the metadata management team and infrastructure specialists. Promoting common use of the metadata
management tools can help influence the positive reception to metadata management and the integration of a metadata-focused approach to information management.
Implement Metadata Management – Each organization will determine their best ways to implement the metadata strategy, standards and tools, but the use of industry experts
often allows the organization to learn from others’ experiences and to focus on the content of the strategy while defining and implementing the recommended policies and procedures.
Metadata management implementation is a lengthy process that will require significant effort from both business and technical staff with the active support of business and technical management.
Following best practices and learning from others can shorten the time to implementation, and can help an organization develop and manage a robust metadata-focused information architecture.