Ecological Metadata Language Overview

1
Ecological Metadata Language Overview

• KNB Data Management Tools Workshop
• Christopher Jones
• Marine Science Institute
• University of California, Santa Barbara

2
Agenda

• Goals of EML
• Module Organization
• Understanding the structure
• Practical Usage
• Development, Documentation and Communication

3
Goals of EML

• Address lack of dataset documentation

4
Goals of EML

• Address lack of dataset documentation
• Provide structure to traditionally unstructured
  information

5
Goals of EML

• Address lack of dataset documentation
• Provide structure to traditionally unstructured
  information
• Aid transition toward synthetic research

6
Goals of EML

• Address lack of dataset documentation
• Provide structure to traditionally unstructured
  information
• Aid transition toward synthetic research
• Enable both human-readable and machine-readable
  metadata

7
Goals of EML

• Address lack of dataset documentation
• Provide structure to traditionally unstructured
  information
• Aid transition toward synthetic research
• Enable both human-readable and machine-readable
  metadata
• Enable long-term archives

8
Goals of EML

• Address lack of dataset documentation
• Provide structure to traditionally unstructured
  information
• Aid transition toward synthetic research
• Enable both human-readable and machine-readable
  metadata
• Enable long-term archives
• Simple transfer format just a text document

9
Goals of EML

• Address lack of dataset documentation
• Provide structure to traditionally unstructured
  information
• Aid transition toward synthetic research
• Enable both human-readable and machine-readable
  metadata
• Enable long-term archives
• Simple transfer format just a text document
• Conform to accepted Internet-based standards for
  the implementation of EML (XML, XML Schema)

10
Goals of EML

• Address lack of dataset documentation
• Provide structure to traditionally unstructured
  information
• Aid transition toward synthetic research
• Enable both human-readable and machine-readable
  metadata
• Enable long-term archives
• Simple transfer format just a text document
• Conform to accepted Internet-based standards for
  the implementation of EML (XML, XML Schema)
• Community-driven development process

11
Goals of EML

• Address lack of dataset documentation
• Provide structure to traditionally unstructured
  information
• Aid transition toward synthetic research
• Enable both human-readable and machine-readable
  metadata
• Enable long-term archives
• Simple transfer format just a text document
• Conform to accepted Internet-based standards for
  the implementation of EML (XML, XML Schema)
• Community-driven development process
• Intended to be both ancillary and integral to
  processing data

12
A simple EML example
13
Can be created in a text editor
lt?xml version"1.0"?gt ltemleml
packageId"sbclter.316.18" system"knb"
xmlnseml"eml//ecoinformatics.org/eml-2.0.1"
xmlnsxsi"http//www.w3.org/2001/XMLSchema-instan
ce" xsischemaLocation"eml//ecoinformatics.org
/eml-2.0.1 eml.xsd" gt ltdatasetgt lttitlegt
Kelp Forest Community Dynamics Benthic Fish
lt/titlegt ltcreatorgt ltindividualNamegt
ltsurNamegtReedlt/surNamegt
lt/individualNamegt lt/creatorgt ltcontactgt
ltindividualNamegt ltsurNamegtEvanslt/surNam
egt lt/individualNamegt lt/contactgt
lt/datasetgt lt/emlemlgt
14
Can be created in custom applications
15
Agenda

• Goals of EML
• Module Organization
• Understanding the structure
• Practical Usage
• Development, Documentation and Communication

16
Module Organization

• Organized as discreet, logical units
• Modular design for re-use of sections
• Self describing field documentation is embedded
  in the EML files themselves

Pete Taylor photo
17
Module Organization (contd)

• Top level eml module is a container, relies on
  sub-modules, also provides identification
  information
• Is extensible in order to accommodate additional
  information

18
Module Organization (contd)

• 4 basic resource types are available
• Each type has unique fields unto itself

19
Module Organization (contd)

• However, they share a set of common fields since
  they may all be considered digital resources
• These align with Dublin Core metadata containers

20
Module Organization (contd)

• Support modules augment each resource with
  broader information
• Provide details for data discovery,
  interpretation, quality

21
Module Organization (contd)

• Dataset organization modules provide details on
  logical structure critical to machine processing
  of data
• Provide schema information objects and
  relationships

22
Module Organization (contd)

• Entities are the detailed data descriptions,
  provide machine-readable details for dataset
  parsing
• Provide syntactic information unique to each data
  type

23
Module Organization (contd)

• Entities are the detailed data descriptions, and
  provide machine-readable details for dataset
  processing
• Provide syntactic information unique to each data
  type

24
Module Organization (contd)

• spatialReference defines fields for coordinate
  systems for referencing spatial coordinates of a
  dataset to the earth (e.g. NAD_1983_UTM_Zone_18N)

25
Module Organization (contd)

• The EML unit dictionary provides pre-defined
  names of units for dataset attributes that are
  mapped back to SI units, with definitions for
  conversion factors, etc.

attribute
26
Module Organization (contd)

• The text module is a utility module that provides
  formatting markup for many fields within EML.
  This allows for structuring paragraphs, bulleted
  lists, etc.

27
Module Organization (contd)

• eml module itself has an extensible section
  called additionalMetadata
• allows for any XML fields to be added to the
  document

additionalMetadata
28
Agenda

• Goals of EML
• Module Organization
• Understanding the structure
• Practical Usage
• Development, Documentation and Communication

29
Top level EML schema structure

• The EML schema is a hierarchical organization
  of the previously discussed modules
• Referred to as a tree of information, with
  sub-trees
• Some elements are optional, some required, etc.

30
Top level EML schema structure
note diagram change
31
Understanding the diagrams
ltemlgt lt/emlgt
element
32
Understanding the diagrams
ltemlgt ltadditionalMetadatagt
lt/additionalMetadatagt lt/emlgt
optional
33
Understanding the diagrams
ltemlgt ltdatasetgt lt/datasetgt
ltadditionalMetadatagt lt/additionalMetadatagt
lt/emlgt
sequence
34
Understanding the diagrams
ltemlgt ltdatasetgt lt/datasetgt
ltadditionalMetadatagt lt/additionalMetadatagt
lt/emlgt
choice
35
Understanding the diagrams
ltemlgt ltdatasetgt lt/datasetgt
ltadditionalMetadatagt lt/additionalMetadatagt
ltadditionalMetadatagt
lt/additionalMetadatagt lt/emlgt
cardinality
36
Understanding the diagrams
ltemlgt ltdatasetgt lt/datasetgt
ltadditionalMetadatagt lt/additionalMetadatagt
ltadditionalMetadatagt
lt/additionalMetadatagt lt/emlgt
complexType
37
Agenda

• Goals of EML
• Module Organization
• Practical Usage
• Specific Features
• Development, Documentation and Communication

38
Practical Usage

• EML is both broad and deep, mostly optional
• Choosing sections to utilize depends on the
  application
• LTER Network is internally looking at best
  practices for providing metadata levels across
  their sites

39
Practical Usage

• EML is both broad and deep, with most fields
  optional
• Choosing sections to utilize depends on the
  application
• LTER Network is internally looking at best
  practices for providing metadata levels across
  their sites

L1
L2
L3
L4
L5
L6
40
Practical Usage

• EML is both broad and deep, with most fields
  optional
• Choosing sections to utilize depends on the
  application
• LTER Network is internally looking at best
  practices for providing metadata levels across
  their sites

T1
T2
T3
T4
T5
T6
During the next session, well explore EML using
the above organization
41
Agenda

• Goals of EML
• Module Organization
• Understanding the structure
• Practical Usage
• Development, Documentation and Communication

42
EML Development Communication

• Open Source project, welcomes contributions
• Developed by members of the community
• eml-dev_at_ecoinformatics.org

Pete Taylor photo

• irc.ecoinformatics.org, eml channel for
  discussion
• Source code managed in cvs.ecoinformatics.org
  repository
• Documented specification found on the KNB website
• An EML validating service is available at
• http//knb.ecoinformatics.org/emlparser

43
EML Distribution

• Currently at Release 2.0.1
• Releases of EML are downloadable
• http//knb.ecoinformatics.org/software/eml
• Development version available at
• http//cvs.ecoinformatics.org/cvs/cvsweb.cgi/e
  ml