The famous Iris dataset exists as a simple CSV everywhere. It’s not FAIR:
Build a proof-of-concept using Airtable (free) + R to transform it into FAIR-compliant JSON-LD.
Three linked tables in Airtable:
R script pulls via API and generates JSON-LD with: - Persistent identifiers for each record - Schema.org vocabulary - Ontology mappings (PO, NCBI) - Full provenance and licensing
library( airtabler )
library( dplyr )
Attaching package: 'dplyr'The following objects are masked from 'package:stats':
filter, lagThe following objects are masked from 'package:base':
intersect, setdiff, setequal, unionlibrary( tidyr )
library(jsonlite)
library(purrr)
Attaching package: 'purrr'The following object is masked from 'package:jsonlite':
flattensource( "./utils.R" )df_at <-
airtable(
base = Sys.getenv("AIRTABLE_BASE") ,
tables = c("Observations", "Property_Mappings (ontology)", "Taxa (species metadata)")
)
df_obs <-
df_at$Observations$select_all() %>%
unnest(c(Species,`Scientific_Name (from Species)`) )
df_ontol <-
df_at$`Property_Mappings (ontology)`$select_all()
df_taxa <- df_at$`Taxa (species metadata)`$select_all()# Config
BASE_DOI <- "https://airtable.com/invite/l?inviteId=invB41lvSB2IoCwtV&inviteToken=912532bce994100be88dfd49cd69d56d0223915e66d6f5b8b431501350901fac&utm_medium=email&utm_source=product_team&utm_content=transactional-alerts"
LICENSE_URL <- "https://creativecommons.org/licenses/by/4.0/"
cat("Converting", nrow(df_obs), "observations to JSON-LD format...\n")Converting 150 observations to JSON-LD format...jsonld_records <- df_obs %>%
rowwise() %>%
do(record = convert_observation_to_jsonld(., df_taxa, df_ontol)) %>%
pull(record)
cat("Conversion complete!\n\n")Conversion complete!dataset_metadata <- list(
`@context` = "http://schema.org",
`@type` = "Dataset",
`@id` = BASE_DOI,
name = "FAIR Iris Dataset",
description = paste(
"Morphological measurements of 150 Iris flowers from three species",
"(Iris setosa, Iris versicolor, and Iris virginica).",
"This dataset demonstrates FAIR data principles with linked ontologies",
"and comprehensive metadata."
),
creator = list(
list(
`@type` = "Person",
name = "Edgar Anderson",
affiliation = "Washington University"
),
list(
`@type` = "Person",
name = "Ronald A. Fisher",
affiliation = "University College London"
)
),
datePublished = "1936",
dateModified = Sys.Date(),
license = LICENSE_URL,
keywords = c("Iris", "morphology", "botany", "classification", "machine learning"),
version = "1.0.0",
distribution = list(
list(
`@type` = "DataDownload",
encodingFormat = "application/ld+json",
contentUrl = paste0(BASE_DOI, "/iris_fair_dataset.json")
)
),
measurementTechnique = "Manual measurement using calipers",
spatialCoverage = "Gaspé Peninsula, Quebec, Canada",
temporalCoverage = "1935",
numberOfRecords = nrow(df_obs),
variableMeasured = df_ontol$Property_Name
)# Print first record nicely formatted
cat(prettify(toJSON(jsonld_records[[1]], auto_unbox = TRUE))){
"@context": "http://schema.org",
"@type": "Dataset",
"@id": "https://airtable.com/invite/l?inviteId=invB41lvSB2IoCwtV&inviteToken=912532bce994100be88dfd49cd69d56d0223915e66d6f5b8b431501350901fac&utm_medium=email&utm_source=product_team&utm_content=transactional-alerts/record/000111",
"identifier": 111,
"name": "Iris specimen 111",
"description": "Morphological measurements of Iris specimen 111",
"license": "https://creativecommons.org/licenses/by/4.0/",
"measurement": [
{
"@type": "PropertyValue",
"propertyID": null,
"name": "sepal length",
"description": null,
"value": 6.5,
"unitCode": null,
"unitText": "centimeters"
},
{
"@type": "PropertyValue",
"propertyID": null,
"name": "sepal width",
"description": null,
"value": 3.2,
"unitCode": null,
"unitText": "centimeters"
},
{
"@type": "PropertyValue",
"propertyID": null,
"name": "petal length",
"description": null,
"value": 5.1,
"unitCode": null,
"unitText": "centimeters"
},
{
"@type": "PropertyValue",
"propertyID": null,
"name": "petal width",
"description": null,
"value": 2,
"unitCode": null,
"unitText": "centimeters"
}
],
"dateCreated": "2026-02-27T08:11:18.000Z",
"about": {
"@type": "Taxon",
"@id": "http://www.ncbi.nlm.nih.gov/Taxonomy/Browser/wwwtax.cgi?id=4449",
"scientificName": "virginica",
"commonName": "Virginia iris",
"taxonRank": "species",
"kingdom": "Plantae",
"family": "Iridaceae",
"genus": "Iris"
},
"provenance": {
"@type": "PropertyValue",
"name": "data provenance",
"collector": {
"@type": "Person",
"name": "Edgar Anderson"
},
"collectionDate": "1935-10-03",
"location": "Gaspé Peninsula, Quebec, Canada"
}
}# Print just the measurement section
cat(prettify(toJSON(jsonld_records[[1]]$measurement, auto_unbox = TRUE)))[
{
"@type": "PropertyValue",
"propertyID": null,
"name": "sepal length",
"description": null,
"value": 6.5,
"unitCode": null,
"unitText": "centimeters"
},
{
"@type": "PropertyValue",
"propertyID": null,
"name": "sepal width",
"description": null,
"value": 3.2,
"unitCode": null,
"unitText": "centimeters"
},
{
"@type": "PropertyValue",
"propertyID": null,
"name": "petal length",
"description": null,
"value": 5.1,
"unitCode": null,
"unitText": "centimeters"
},
{
"@type": "PropertyValue",
"propertyID": null,
"name": "petal width",
"description": null,
"value": 2,
"unitCode": null,
"unitText": "centimeters"
}
]# Print dataset metadata
cat(prettify(toJSON(dataset_metadata, auto_unbox = TRUE))){
"@context": "http://schema.org",
"@type": "Dataset",
"@id": "https://airtable.com/invite/l?inviteId=invB41lvSB2IoCwtV&inviteToken=912532bce994100be88dfd49cd69d56d0223915e66d6f5b8b431501350901fac&utm_medium=email&utm_source=product_team&utm_content=transactional-alerts",
"name": "FAIR Iris Dataset",
"description": "Morphological measurements of 150 Iris flowers from three species (Iris setosa, Iris versicolor, and Iris virginica). This dataset demonstrates FAIR data principles with linked ontologies and comprehensive metadata.",
"creator": [
{
"@type": "Person",
"name": "Edgar Anderson",
"affiliation": "Washington University"
},
{
"@type": "Person",
"name": "Ronald A. Fisher",
"affiliation": "University College London"
}
],
"datePublished": "1936",
"dateModified": "2026-02-27",
"license": "https://creativecommons.org/licenses/by/4.0/",
"keywords": [
"Iris",
"morphology",
"botany",
"classification",
"machine learning"
],
"version": "1.0.0",
"distribution": [
{
"@type": "DataDownload",
"encodingFormat": "application/ld+json",
"contentUrl": "https://airtable.com/invite/l?inviteId=invB41lvSB2IoCwtV&inviteToken=912532bce994100be88dfd49cd69d56d0223915e66d6f5b8b431501350901fac&utm_medium=email&utm_source=product_team&utm_content=transactional-alerts/iris_fair_dataset.json"
}
],
"measurementTechnique": "Manual measurement using calipers",
"spatialCoverage": "Gaspé Peninsula, Quebec, Canada",
"temporalCoverage": "1935",
"numberOfRecords": 150,
"variableMeasured": [
"sepal_length ",
"sepal_width ",
"petal_length ",
"petal_width "
]
}# Export individual records
cat("Exporting JSON-LD files...\n")Exporting JSON-LD files...# Create output directory
dir.create("fair_iris_output", showWarnings = FALSE)
dir.create("fair_iris_output/records", showWarnings = FALSE)
# Export each record individually
for (i in seq_along(jsonld_records)) {
record_id <- sprintf("%06d", df_obs$Record_ID[i])
filename <- paste0("fair_iris_output/records/iris_record_", record_id, ".json")
write(
toJSON(jsonld_records[[i]], pretty = TRUE, auto_unbox = TRUE, null = "null"),
file = filename
)
}
cat(" ✓ Exported", length(jsonld_records), "individual record files\n") ✓ Exported 150 individual record files# Export complete dataset (all records in one file)
complete_dataset <- list(
`@context` = "http://schema.org",
`@graph` = jsonld_records
)
write(
toJSON(complete_dataset, pretty = TRUE, auto_unbox = TRUE, null = "null"),
file = "fair_iris_output/iris_complete_dataset.json"
)
cat(" ✓ Exported complete dataset file\n") ✓ Exported complete dataset file# Export dataset metadata
write(
toJSON(dataset_metadata, pretty = TRUE, auto_unbox = TRUE, null = "null"),
file = "fair_iris_output/iris_dataset_metadata.json"
)
cat(" ✓ Exported dataset metadata file\n") ✓ Exported dataset metadata file