EML

EML Best Practices for LTER Sites

EML Best Practices, V2
August, 2011:
Version 2 of "EML Best Practices for LTER Sites" is now available, and the pdf is attached below with examples. The working group responsible can be found at:
Projects > EML Best Practices

Earlier versions of the document are available from the network document archive.

EML Handbook, 2003

II. Detailed content recommendations For Elements and Attributes

Following are general best practices for creating EML dataset metadata:

Metadata Distribution: Do not publicly distribute EML documents containing elements with incorrect information as dataset metadata (i.e. as a workaround for problems with metadata content availability or to meet EML validation requirements). EML produced as draft, demonstration or for testing purposes should be clearly identified as such and not contributed to public metadata archives or clearinghouses.

I.2.3. EML produced from Geographic Information Systems (GIS) systems

Several established standards exist for documenting spatial datasets. The most common formats are Federal Geographic Data Committee (FGDC) geospatial standard, National Biological Information Infrastructure (NBII) biological profile, International Standards Organization (ISO) standard 19115, and ArcGIS metadata format. Tools are available at http://www.fgdc.gov/metadata/geospatial-metadata-tools for preparing FGDC compliant metadata, which can then be converted into EML using XSL transformation style sheets.

I.2.2 The Attribute – Value Data Model

The Attribute - Value or “string of pearls” data model is widely used for certain kinds of observational data where the more conventional matrix type model would cause many empty cells.

I.2.1. Creating Datasets

Several approaches to creating datasets or data packages have emerged in the network and are all valid concepts. In general, this document does not recommend any one pattern. Following are several examples:
(1) Data collected with defined beginning and end dates are published in logical units, with all ancillary data are described together in one EML file (i.e., <dataset>). EML accommodates descriptions of multiple data entities in one metadata document. For instance, a data table could be accompanied by a KML file or shapefile to describe the sampling locations.

Syndicate content