Information systems rely on metadata to manage information. Metadata is the data that describes data or information.
Records metadata is divided into three main types:
- structural metadata – used to describe the relationship between records, such as a file or container, or the format of the record
- administrative metadata – used to manage a resource, such as access permissions, and
- descriptive metadata.
Descriptive metadata is deliberate, structured data that provides information about the content or context of the record which may not otherwise be recognised by full-text searching.
Descriptive metadata includes:
- unique identifier
- title
- subjects
- author
- document type
- description
- keywords
There is a direct relationship between the metadata assigned to a record and the discoverability of that record.
Metadata is generally defined in a schema and encoding scheme or controlled vocabulary.
- A metadata schema shows the relationship between metadata elements and establishes rule for use, e.g. is the element optional or mandatory, which element takes precedence.
- An encoding scheme is the controlled list of terms used and includes rules for how the terms should be entered, e.g. Surname, First name, date formats, etc.
What Metadata do I need?
Requirements for the use of metadata depends on the needs of the users and the functionality of the system being used. Does the system:
- allow for full text searching?
- use containers?
- automatically generate metadata through the capture process?
- use templates that have specific metadata imbedded, e.g. a Leave Application or Accident Report Form
- use artificial intelligence or has machine learning functionality.
Automating the capture of metadata where possible reduces the need for manually entering information and enables more consistent classification.
When deciding what metadata to capture you should also consider the value of the records. Are they:
- of long term value to the organisation
- of archival value
- vital for the continuation of business functions.
The more records you have in a collection the more critical the metadata becomes.
Managing metadata
Metadata should never be created on an ad-hoc basis, it should be managed.
Properly managed metadata promotes the entry of meaningful, standardised and consistent metadata by the users. This contributes to record quality, and discoverability, assists in record sharing and reuse, and increases the efficiency of machine learning where available.
Like all records management functions, metadata and its use should be documented. This information is very important for ongoing systems management and for facilitating system migration in the future.
As an organisation’s business activities change, the metadata schemas and encoding schemes should be reviewed to make sure they still meet the organisation’s needs.
