lab

 

JSON-LD (JavaScript Object Notation for Linked Data) is a recent recommendation from the W3C.  The goal of JSON-LD was to take the compact JSON format and use it to store metadata in a way that Resource Description Framework (RDF) triples could be generated directly from JSON-LD files. As it is important to show the versatility of ChAMP, in this case across different file formats, a JSON-LD context for ChAMP was developed in order to create example standards from it.

The main context file and the contexts referenced within are shown below (with links to the files).  Unlike XML where schema are developed as single files (in most cases) due to the organized structure of XML documents, JSON-LD is setup to allow arrays of contexts (and additional definitions) be integrated together (by the order they are in the array). Below are graphical representations of the main categories of metadata (each as a context) plus some additional metadata elements that complete the ChAMP specification (the JSON has be compacted a little to make the page a little shorter).

Main Context

main

Analyte Context

analyte

Concept Context

concept

Contact Context

contact

Data Context

data

Datum Context

datum

Description Context

description

Infrastructure Context

infrastructure

Materials Context

materials

Metadata Context

metadata

Quality Context

quality

Sample Context

sample

Sampleprep Context

sampleprep

 Solution Context

solution

Substance Context

substance

Technique Context

technique

This example is an analogous standard to the XML Research Article Metadata standard, except in the JSON-LD format.  Starting from the ChAMP JSON-LD context a research article context was developed to mirror the metadata in the XML version.  Note below that unlike the XML version the context only needs to define those name/value pairs that are not already defined in the ChAMP context.  This is because JSON-LD is more free form than XML and does not require elements to be in any particular order - the reason the W3C committee did not call the context a 'schema'.

article context

Looking at the equivalent example article file below we see the cleaner nature of the JSON format where {} define objects and [] define arrays.  This file can be downloaded here.  To convert the file to different JSON-LD formats and RDF visit the JSON-LD Playground (NOTE:  the contexts for the standard are available at https://chalk.coas.unf.edu so that they work correctly with the JSON-LD playground).

Example File - DOI:10.1021/ac503447m

article jsonld

Although ChAMP is a platform of metadata that is not tied to any one format for implementation, in terms of translating to potential users ChAMPs useablility, it is necessary to create some example applications. In thinking about these applications it became obvious that an additional resource the project should provide was a schema for XML based implementations of the platform (schema equivalents for other formats are in development).

The schema can be downloaded here.  Below are graphical representations of the main categories of metadata were identified as being important to characterize a chemical analysis (sorry for the size of the page).

 champ schema analyte

champ schema concept

 champ schema description

champ schema infrastructure

champ schema materials

champ schema metadata

champ schema quality

champ schema sample

champ schema sampleprep

champ schema technique

As a first example of the use of ChAMP in creation of a standard, research article metadata was chosen.  When we research the literature, we see that publishers provide keywords with the bibliographic information and the abstract on the page for the article.  While the keywords are important - they are what the papers' author(s) think are important facets of the research described - they are not standardized and thus do not really fit the needs of the readers of the paper.

In most cases, we find a paper via a text search of the title and abstract and then look for specific metadata (detection limit, analysis time, instrumentation, method accuracy, etc.) to evaluate whether it is something we can use, compare our work to, etc.  With that in mind, using ChAMP to represent important chemical analysis metadata (that researchers would like to search and compare methods with) is a logical use of the platform.

Starting from the ChAMP XML Schema any/all pieces of the schema can be used to create a schema for this application.  Below is an image of the schema for this example.  Remember, standards authors can design this any way they think is appropriate for their application.

[Other links:  ChAMP Standard Method XML ExampleChAMP Referemce Material XML Example]

 article schema

As you can see, the schema picks only certain elements for the standard.  It shows how to re-purpose existing elements ('champ:description') with a new name ('overview') using the 'substitutionGroup' attribute.  You can also import other schema, if you want a standard to use elements from schema other that ChAMP. Below is what an example XML file might look like representing the metadata about a specific paper.

article xml

Alternatively, the links to the ChAMP ontology can be added to the XML file as seen below

article xml ont

Either way, because the ChAMP XML schema is designed with the ontology references included, the metadata items can be crosswalked to the ontology by using the schema.