• Understand the concept of metadata

• Understand the different kinds of metadata

• Get an introduction to the Ecological Metadata Language

“data that provides information about other data”

“shorthand representation of the data to which they refer”

• title
  
• author
  
• publisher
  
• copyright
  
• table of contents / index

• number called or received from
  
• time initiated & ended
  
• location of callers

• subject
  
• from
  
• to
  
• date & time sent
  
• sending & receiving IP addresses

Metadata benefits science by

1. increased data longevity

   • long-term studies outlast investigators/students
   • many scientists contribute info from many areas

Metadata benefits science by

1. increased data longevity

2. increased data reuse & sharing

   • info helps others know if/how they could them
   • facilitates meta-analyses

Metadata benefits science by

1. increased data longevity

2. increased data reuse & sharing

3. expanded scales/scopes of analyses

   • short-term evolve into long-term
   • facilitates creativity

Descriptive metadata is information about a resource that is used for searching & identification

• title
  
• abstract
  
• author
  
• keywords

Structural metadata indicates how 2+ objects are connected

• tables
  
• fields
• keys
  
• relationships

Structural metadata are often conceptualized in an entity–relationship model

An entity can be physical or logical

• physical entities actually exist (eg, lake, fish)

• logical entities could exist (eg, a sampling event)

• entities have attributes (eg, depth, mass, date)

Entities are connected via a relationship

• relationships can be thought of as verbs

• Sarah supervises Mark

• Omar samples Lake Washington

ER models are usually drawn as boxes (entities) connected by lines (relationships)

Administrative metadata refers to technical info about a data file

• file type (eg, .csv)
  
• when the file was created
• how the file was created

• Who has had custody/ownership of the data?
  
• What intellectual property rights must be observed?

• Where are the data stored?
  
• How are they identified (eg, digitial object identifier)?

Information about the contents and quality of statistical data

• description of fields
  
• QA/QC

Describes the collection and any processing of the data

• raw vs cleaned
• reproducible workflows

Information about improved access to data

• audio transcripts
  
• alternate text for an image
  
• large print

Two factors to consider

1. effort involved in creating the metadata

2. value derived from it

In general, assume that “more is better”

EML describes a range of essential aspects of ecological data

• names & definitions of variables
  
• units of measurement
  
• date/time/location of data collection
  
• identity/contact for individual who collected the data
  
• sampling design

EML reduces ambiguity & uncertainty by formalizing metadata concepts

• uses comprehensive and standardized set of terms
  
• uses definitions intended specifically for ecological data

1. General dataset

2. Geographic

3. Temporal

4. Taxonomic

5. Methods

6. Data table

Contains concepts that

• identify and name the dataset
  
• describe the purpose of the data collection
  
• describe the questions the data were intended to address

Contains information about where

• the research project took place
  
• the samples were collected
  
• spatial or geographic references (UTM, lat/lon)

Contains information about when

• range of dates (monthly between June 2019 and Dec 2020)
  
• specific time periods (01 May 2019, 08:00–12:00)
  
• gaps in time (no data from July 2020 because of power loss)

Contains information about

• taxonomic authority (book or system used to identify species)
  
• taxonomic rank (family, genus, species)

Contains information about what happened

• instruments or devices used to collect data
  
• protocols
• units of the samples

Contains information about what happened

• instruments or devices used to collect data
  
• protocols
• units of the samples

Unlike the Methods section of a publication, fully detailed descriptions can be included

Contains information about a rectangular table

• file name
  
• number of records
• structure of the table (attributes of fields/columns)

• name is a unique name for the field/column (date)
  
• label describes the field/column ("date of sample collection)
  
• definition indicates what the values represent (length of a fish)
  
• units (grams, meters)
  
• type (numeric, character)
  
• precision (mm)
  
• attribute domain description defines codes & domain of values
• BVA = Bear Valley Creek
  
• Length is a positive, real value

The KNB is an international repository intended to facilitate ecological & environmental research

• you can upload/publish data with a DOI

• you can write EML

Provides read & write access to data and metadata from DataONE network

• KNB Data Repository
  
• Dryad
  
• NEON
• USGS
• NOAA

## username
me <- list(individualName = list(givenName = "Mark",
                                 surName = "Scheuerell"))

## list of attributes
my_eml <- list(
  dataset = list(
    title = "A Minimal Valid EML Dataset",
    creator = me,
    contact = me
  )
)

## inspect the EML
my_eml

We’ll discuss data analysis and visualization