Metadata Repositories
Contents |
CTRMeta and CTRWiki
CTRNet has developed two web applications based on the ISO standard for metadata repositories.
- CTRMeta - Built on the ATiM Core framework. Used by librarians and group librarians to create and annotate all items types (properties, objects, data elements, etc) within the MDR.
- CTRWiki - Public interface to the metadata repository. Used for easy browsing of existing elements and output of data to PDF for distribution to work groups for review.
Metadata Registry
A Metadata Registy or Repository is a database of metadata that supports the functionality of registration. An MDR manages the semantics of data. Registration accomplishes three main goals: identification, provenance and monitoring quality.
Identification – Assigning unique ID within registry to each object registered Provenance – addresses the source of the metadata and the object described Monitoring – Ensures metadata does the job it is designed to do
In order to discuss a Metadata Repository/Registry (MDR) some terminology needs to be defined. These terms and definitions are common among other MDR's following the ISO/IEC 11179 Standard.
About metadata standards
- The development of metadata standards improves the quality, relevance, consistency and availability of information on cancer and tumour research at a national level. The Standard development of metadata definitions arises from the need for better information - whether it is statistical, administrative, clinical or otherwise.
- Metadata standards describe the expected meaning and acceptable representation of data for use within a defined context. The need for consistency of meaning is vital to facilitate information sharing among primary and secondary users of the data. Much of the work involved in establishing a data collection lies in the development of metadata standards to ensure comparability and consistency of the data collected.
- If we never have to share data then there is no need to standardize. If we share data then we need to ensure that all those who need to use the data can clearly understand the meaning regardless of how the data is collected or stored.
- Avoid duplication and diversity of solutions: Metadata standards are generally required when excessive diversity creates inefficiencies or impedes effectiveness. Metadata standards offer a means of narrowing the variety of ways information is exchanged among different groups, allowing synergy between multiple development efforts.
- Reduction in cost of data development: Metadata standards provide a way to solve a problem that other people can use without having to start from scratch. Metadata standards provide a common and consistent platform for organizations to work, thereby simplifying adoption and implementation at the local and national levels.
Building blocks of metadata
- Dataset - A collection of data elements.
- Data Element – Basic container for data. Descriptions have both semantic and representational components. The semantics are further divided into contextual and symbolic types.
- Data Element Concept – a concept that can be represented in the form of a data element, described independently of any particular representation.
- Object – anything perceivable or conceivable. Objects may also be material (engine, paper, diamond), immaterial (conversion ratio, project plan) or imagined (dragon, Bigfoot).
- Property – characteristic common to all members of an object class.
- Value Domain – Specifies all the values that are allowed either through an enumeration, a rule, or a combination of these. The computational model the values follow is given by their datatype.
- Representation – expression of a value meaning allowed in a specific value domain.
- Concept – mental constructs, units of thought, or unit of knowledge created by a unique combination of characteristics. Organized or grouped by common elements, called characteristics.
- Classification Scheme – the descriptive information for an arrangement or division of objects into groups based on characteristics, which objects have in common.
- Conceptual Domain - A set of valid value meanings, which may be enumerated or expressed via a description.
- Value Meaning - The meaning or semantic content of a value.
- Qualifier - A word or words that help define and differentiate a name within a context.
- Specialisation - Specialisations allow object classes to be grouped and sub-typed in a meaningful manner and help users in browsing and locating relevant object classes. Sub-typed objects inherit the properties of the generalised object, in additional to having their own unique properties.
Language and spelling
Use British English spelling unless referring to an organization which applies another language in its formal title or when referring to a “foetus”. In these cases, state the formal organization title, e.g., World Health Organization, and use the British English spelling — ‘fetus’.
Abbreviation
Express the first occurrence of an abbreviation within each metadata item by spelling out the word or phrase in full and enclosing the abbreviation within parentheses immediately to the right of this word. For example, The International Statistical Classification of Diseases and Related Health Problems, Tenth Revision, Australian Modification, 6th edition (ICD-10-AM 6th edn.). This rule does not apply to the abbreviated classification scheme name included in value domain and data element names.
