What is it?
One definition of metadata often used is that metadata is "Data about other data"
This is not particularly helpful to researchers, librarians, or research administrators who are approaching research data management for the first time.
A more useful definition, in this context is:
Metadata is a description of the data in order to help people to obtain and use the data.
This description can be background information on the data (e.g. "This data originated as records of births in the country"), as well as information on how the data can be used (e.g. "This data is not reliable at district level but can be used to examine the phenomenon at country level").
Standards for creating metadata
The data curation community has developed standards to assist those creating these descriptions and ensure that their metadata is structured enough to be machine-readable to be converted to other formats.
There are standards for describing whole datasets e.g. the Data Documentation Initiative metadata standard. There are also standards for describing entities within a dataset e.g. the Dublin Core metadata standard, which uses the Resource Description Framework to describe documents and data files within a dataset.
Software for creating metadata
Metadata creation software provides ready-made fields for data descriptions, which can be customised and produce xml compliant metadata which can be displayed online.
e.g. Nesstar Publisher, which is free data mark-up software for the creation of XML compliant metadata. This software uses the Data Documentation Initiative (DDI) metadata standard for describing whole datasets, and the Dublin Core standard for describing the documents and data files within a dataset.