This document defines the procedures and rules to be applied when converting tabular data into RDF. Tabular data may be complemented with metadata annotations that describe its structure, the meaning of its content and how it may form part of a collection of interrelated tabular data. This document specifies the effect of this metadata on the resulting RDF.

The CSV on the Web Working Group was chartered to produce a recommendation "Access methods for CSV Metadata" as well as recommendations for "Metadata vocabulary for CSV data" and "Mapping mechanism to transforming CSV into various formats (e.g., RDF, JSON, or XML)". This document aims to satisfy the RDF variant of the mapping recommendation.

Introduction

This document describes the processing of tabular data to create an RDF subject-predicate-object triples [[!rdf11-concepts]]. Since RDF is an abstract syntax, these triples MAY be serialized in a concrete RDF syntax such as N-Triples [[n-triples]], Turtle [[turtle]], RDFa [[rdfa-primer]], JSON-LD [[json-ld]], or TriG [[trig]]. The RDF serializations offered by a conversion application is implementation defined.

The [[!tabular-data-model]] defines an annotated tabular data model consisting of tables, columns, rows, and cells, enriched with annotations that describe the structure of the tabular data and the meaning of its content. A group of tables is a collection of tables published as a single atomic unit.

The conversion procedure described in this specification operates on the annotated tabular data model. This specification does not specify the processes needed to convert CSV-encoded data into tabular data form. Please refer to [[!tabular-data-model]] for details of parsing tabular data.

Conversion applications MUST provide at least two modes of operation: standard and minimal.

Standard mode conversion frames the information gleaned from the cells of the tabular data with details of the rows, tables, and a group of tables within which that information is provided.

Minimal mode conversion includes only the information gleaned from the cells of the tabular data.

Standard and minimal conversion are described normatively below.

Conversion applications MAY offer additional implementation specific conversion modes.

Transformation definitions, as defined in [[!tabular-metadata]] MAY be used to specify how tabular data can be transformed into another format using a script or template. Such transformation definitions MAY use the RDF output described in this specification as input.

There is no requirement on conversion applications to check the semantic consistency of the data during the conversion, nor to validate the triples against RDF schema. Downstream applications SHOULD be aware of the potential for inconsistencies and take appropriate action.

Tabular data MUST conform to the description from [[!tabular-data-model]]. In particular note that each row MUST contain the same number of cells (although some of these cells may be empty).

Not all CSV-encoded data can be parsed into a tabular data model. An algorithm for parsing CSV-based files is described in [[!tabular-data-model]].

This specification makes use of the compact IRI Syntax; please refer to the Compact IRIs from [[json-ld]].

This specification makes use of the following namespaces:

csvw:
http://www.w3.org/ns/csvw#
rdf:
http://www.w3.org/1999/02/22-rdf-syntax-ns#
xsd:
http://www.w3.org/2001/XMLSchema#

Typographical conventions

Converting Tabular Data to RDF

The procedures for converting tabular data into RDF are described below for both standard and minimal modes.

Algorithm terms

about URL
The about URL annotation on the current cell. As defined in [[!tabular-data-model]].
annotated table
The annotated table is defined in [[!tabular-data-model]] as describing a particular table and its annotations.
blank node
A blank node is defined in [[!rdf11-concepts]] as an RDF Term disjoint from IRIs or literals.
cell
A cell is defined in [[!tabular-data-model]] as the intersection of a row and a column within a table.
cell errors
Cell errors are defined in [[!tabular-data-model]] as a (possibly empty) list of validation errors generated while parsing the literal content of a cell to generate the semantic value.
cell value
A cell value is defined in [[!tabular-data-model]] as the semantic value of the cell; this MAY be null or a sequence of values.
column
A column is defined in [[!tabular-data-model]] as a vertical arrangement of cells within a table.
group of tables
A group of tables is defined in [[!tabular-data-model]] as comprising a set of annotated tables and a set of annotations that relate to that group.
group of tables identifier
The group of tables identifier is the id annotation on a group of tables. As defined in [[!tabular-data-model]].
literal node
A literal node is defined in [[!rdf11-concepts]] as a node within an RDF graph that provides values such as strings, numbers, and dates.
node
A node is defined in [[!rdf11-concepts]] as a subject or an object of an RDF triple. When in subject position, it can be either a blank node or identified with a URL; when in object position, it can be a blank node, a literal, or identified with a URL.
non-core annotations
Core annotations are listed in [[!tabular-data-model]]; groups of tables and tables may also have other annotations that are not defined in that specification; these are known as non-core annotations.
notes
A list of notes, as defined in [[!tabular-data-model]], attached to an annotated table or group of tables using the notes property. This may be an empty list.
predicate
A predicate is defined in [[!rdf11-concepts]] as an IRI that denotes the property used to relate nodes within an RDF triple.
prefixed name
A prefixed name is an abbreviation for a URI, in the syntax prefix:name. See Names of Common Properties in [[!tabular-metadata]] for information on expansion.
property URL
The property URL annotation on the current cell. As defined in [[!tabular-data-model]].
row
The row is defined in [[!tabular-data-model]] as a horizontal arrangement of cells within a table.
row number
A row number is defined in [[!tabular-data-model]] as the position of the row within the table, starting from 1.
row source number
A row source number is defined in [[!tabular-data-model]] as the position of the row within the source tabular data file. Provision of the row source number is dependent on parsing applications and may be reported as null.
subject
Within this algorithm, a subject is the resource that the value of a given cell refers to. This may be specified using about URL.
table identifier
The table identifier is the id annotation on an annotated table. As defined in [[!tabular-data-model]].
tabular data mapping
The mapping from tabular data to RDF, as defined by this Recommendation.
value URL
The value URL annotation on the current cell. As defined in [[!tabular-data-model]].

Generating RDF

A conformant RDF conversion application MUST emit triples conforming to those described in this algorithm according to the chosen mode of conversion: standard or minimal.

Unless specified otherwise, the steps in the algorithm defined herein apply to both standard and minimal modes.

Where an annotated table is defined in isolation (e.g. in the absence of a group of tables), a default group of tables is provided with a single tables annotation that refers to the given table.

The [[!tabular-data-model]] specifies that string values within tabular data (such as column titles or cell string values) MUST contain only Unicode characters. No Unicode normalization (as specified in [[UAX15]]) is applied to these string values during the conversion to RDF.

If a CSV file is originally encoded as UTF-8, it should not go through Unicode normalization during parsing, nor in conversion to RDF. This can result in RDF literals that are not in Normal Form C as they should be according to [[rdf11-concepts]].

  1. In standard mode only, establish a new node G. If the group of tables has an identifier then node G MUST be identified accordingly; else if identifier is null, then node G MUST be a new blank node.
  2. In standard mode only, specify the type of node G as csvw:TableGroup; emit the following triple:

    subject
    node G
    predicate
    rdf:type
    object
    csvw:TableGroup
  3. In standard mode only, emit the triples generated by running the algorithm specified in over any notes and non-core annotations specified for the group of tables, with node G as an initial subject, the notes or non-core annotation as property, and the value of the notes or non-core annotation as value.

  4. For each table where the suppress output annotation is false:

    1. In standard mode only, establish a new node T which represents the current table.

      If the table has an identifier then node T MUST be identified accordingly; else if identifier is null, then node T MUST be a new blank node.

    2. In standard mode only, relate the table to the group of tables; emit the following triple:

      subject
      node G
      predicate
      csvw:table
      object
      node T
    3. In standard mode only, specify the type of node T as csvw:Table; emit the following triple:

      subject
      node T
      predicate
      rdf:type
      object
      csvw:Table
    4. In standard mode only, specify the source tabular data file URL for the current table based on the url annotation; emit the following triple:

      subject
      node T
      predicate
      csvw:url
      object
      a node identified by URL
    5. In standard mode only, emit the triples generated by running the algorithm specified in over any notes and non-core annotations specified for the table, with node T as an initial subject, the notes or non-core annotation as property, and the value of the notes or non-core annotation as value.

      All other core annotations for the table are ignored during the conversion; including information about table schemas and columns specified therein, foreign keys, table direction, transformations etc.

    6. For each row in the current table:

      1. In standard mode only, establish a new blank node R which represents the current row.

      2. In standard mode only, relate the row to the table; emit the following triple:

        subject
        node T
        predicate
        csvw:row
        object
        node R
      3. In standard mode only, specify the type of node R as csvw:Row; emit the following triple:

        subject
        node R
        predicate
        rdf:type
        object
        csvw:Row
      4. In standard mode only, specify the row number n for the row; emit the following triple:

        subject
        node R
        predicate
        csvw:rownum
        object
        a literal n; specified with datatype IRI xsd:integer
      5. In standard mode only, specify the row source number nsource for the row within the source tabular data file URL using a fragment-identifier as specified in [[RFC7111]]; if row source number is not null, emit the following triple:

        subject
        node R
        predicate
        csvw:url
        object
        a node identified by URL#row=nsource
      6. In standard mode only, if row titles is not null, insert any titles specified for the row. For each value, tv, of the row titles annotation, emit the following triple:

        subject
        node R
        predicate
        csvw:title
        object
        a literal tv; specified with the the appropriate language tag (as defined in [[!rdf11-concepts]]) for that row title annotation value
      7. In standard mode only, emit the triples generated by running the algorithm specified in over any non-core annotations specified for the row, with node R as an initial subject, the non-core annotation as property, and the value of the non-core annotation as value.

      8. Establish a new blank node Sdef to be used as the default subject for cells where about URL is undefined.

        A row MAY describe multiple interrelated subjects; where the value URL annotation on one cell matches the about URL annotation on another cell in the same row.

        For each cell in the current row where the suppress output annotation for the column associated with that cell is false:

        1. Establish a node S from about URL if set, or from Sdef otherwise as the current subject.

        2. In standard mode only, relate the current subject to the current row; emit the following triple:

          subject
          node R
          predicate
          csvw:describes
          object
          node S
        3. If the value of property URL for the cell is not null, then predicate P takes the value of property URL.

          Else, predicate P is constructed by appending the value of the name annotation for the column associated with the cell to the the tabular data file URL as a fragment identifier.

        4. If the value URL for the current cell is not null, then value URL identifies a node Vurl that is related the current subject using the predicate P; emit the following triple:
          subject
          node S
          predicate
          P
          object
          node Vurl
        5. Else, if the cell value is a list and the cell ordered annotation is true, then the cell value provides an ordered sequence of literal nodes for inclusion within the RDF output using an instance of rdf:List Vlist as defined in [[rdf-schema]]. This instance is related to the subject using the predicate P; emit the triples defining list Vlist plus the following triple:
          subject
          node S
          predicate
          P
          object
          node Vlist
        6. Else, if the cell value is a list, then the cell value provides an unordered sequence of literal nodes for inclusion within the RDF output, each of which is related to the subject using the predicate P. For each value provided in the sequence, add a literal node Vliteral; emit the following triple:
          subject
          node S
          predicate
          P
          object
          literal node Vliteral
        7. Else, if the cell value is not null, then the cell value provides a single literal node Vliteral for inclusion within the RDF output that is related the current subject using the predicate P; emit the following triple:
          subject
          node S
          predicate
          P
          object
          literal node Vliteral

          The literal nodes derived from the cell values MUST be expressed according to the cell value's datatype as defined below: Interpreting datatypes.

          In the case when a cell value does not have a datatype, the conversion should default to string.

          In the case where a sequence of values is provided, each value in the list has its own datatype; the datatype may be different for different items in the sequence.

Interpreting datatypes

Cell values are expressed in the RDF output according to the cell value's datatype. The relationship between the value of the cell value's datatype and the datatype IRI used in the RDF output is as follows:

A datatype's format annotation is irrelevant to the conversion procedure defined in this specification; the cell value has already been parsed from the contents the cell according to the format annotation.

Where the contents of the cell cannot be parsed, or other validation errors occur, cell errors will be provided. It is an implementation decision to determine how conversion applications should proceed in the event that cell errors are encountered.

datatypeRDF datatype IRIRemarks
anyAtomicTypexsd:anyAtomicType
anyURIxsd:anyURI
base64Binaryxsd:base64Binary
booleanxsd:boolean
datexsd:date
dateTimexsd:dateTime
dateTimeStampxsd:dateTimeStamp
decimalxsd:decimal
integerxsd:integer
longxsd:long
intxsd:int
shortxsd:short
bytexsd:byte
nonNegativeIntegerxsd:nonNegativeInteger
positiveIntegerxsd:positiveInteger
unsignedLongxsd:unsignedLong
unsignedIntxsd:unsignedInt
unsignedShortxsd:unsignedShort
unsignedBytexsd:unsignedByte
nonPositiveIntegerxsd:nonPositiveInteger
negativeIntegerxsd:negativeInteger
doublexsd:double
durationxsd:duration
dayTimeDurationxsd:dayTimeDuration
yearMonthDurationxsd:yearMonthDuration
floatxsd:float
gDayxsd:gDay
gMonthxsd:gMonth
gMonthDayxsd:gMonthDay
gYearxsd:gYear
gYearMonthxsd:gYearMonth
hexBinaryxsd:hexBinary
QNamexsd:QName
stringxsd:string or rdf:langString depending on whether or not the value has an associated language.
normalizedStringxsd:normalizedString
tokenxsd:token
languagexsd:language
Namexsd:Name
NMTOKENxsd:NMTOKEN
xmlrdf:XMLLiteral
htmlrdf:HTML
jsoncsvw:JSONcsvw:JSON is a sub-class of xsd:string
timexsd:time

In the case of rdf:langString, the appropriate language tag (as defined in [[!rdf11-concepts]]) MUST be provided for the string, based on the value of cell value's language. (See section on Graph Literals in [[!rdf11-concepts]] for further details on language tagged literals.)

According to [[rdf11-concepts]] language tags cannot be combined with any other xsd datatypes. If a cell has any other datatype than string, the value of lang MUST be ignored. Also, all literals have a datatype; however, specific serializations, like Turtle [[turtle]], MAY provide a special syntax for literals with datatype xsd:string or rdf:langString.

Inclusion of provenance information

In addition to the namespaces defined above, the following namespace is used in this section:

prov:
http://www.w3.org/ns/prov#

Conversion applications MAY include provenance information in the RDF output describing how and when the output was created; e.g., using terms from the PROV Ontology [[prov-o]]. Information that may be of interest to downstream applications includes:

In order to faciliate the provision of such information, this specification introduces two instances of prov:Role:

csvw:csvEncodedTabularData
Defines the role of the source tabular data file.
csvw:tabularMetadata
Defines the role of the metadata description file.
An illustrative example of provenance information is provided below in Turtle [[turtle]] syntax, the conversion application used is identified as http://example.org/my-csv2rdf-application:
       

JSON-LD to RDF

This section defines a mechanism for transforming the [[json-ld]] dialect used for non-core annotations and notes originating from the processing of metadata (as defined in [[!tabular-metadata]]) into RDF in a manner consistent with the Deserialize JSON-LD to RDF Algorithm defined in [[!json-ld-api]]. Converters MAY use any algorithm which results in equivalent triples.

Conversion applications may have other means to create annotated tables, e.g., through some application specific API-s. In such cases the exact format for non-core annotations or notes may be different. Specifications for such annotation processes should specify how these annotations should be converted into RDF.

Given a subject, property and value in normalized form:

  1. Property is a term defined in the [[csvw-context]], a prefixed name, or an absolute URL; expand to an absolute URL by replacing a term with the URI from the term definition in [[csvw-context]] or a prefixed name as described in Names of Common Properties in [[!tabular-metadata]].
  2. If value is an array, generate RDF by running this algorithm using subject, property using each array member as value.
  3. If value is an object containing @value, create an RDF Literal lit using the string value of @value and language from @language, or datatype from @type if present, expanding @type as necessary using the procedure outlined for property, and emit the following triple:
    subject
    node subject
    predicate
    property
    object
    literal node lit

    If neither @language nor @type is present, the literal lit has the datatype xsd:string.

  4. Else, if value is an object:
    1. Establish a new node S from the value of @id, if it exists, and new blank node otherwise and emit the following triple:
      subject
      node subject
      predicate
      node property
      object
      node S
    2. For every value of @type, either a term defined in the [[csvw-context]], a prefixed name, or an absolute URL; establish a new node Ti by expanding the value to an absolute URL by replacing a term with the URI from the term definition in [[csvw-context]] or a prefixed name with its expanded value. For each Ti, emit the following triple:
      subject
      node S
      predicate
      rdf:type
      object
      node Ti
    3. For every key and val from value that does not start with @ (U+0040) generate RDF by running this algorithm using S for subject, key for property and val for value.
  5. Else, establish lit as an RDF Literal as follows:
    1. If value is true or false, create an RDF Literal lit using the strings "true" or "false", accordingly with datatype xsd:boolean
    2. Else, if value is a JSON number with a non-zero fractional part, create an RDF Literal lit using the canonical representation for value with datatype xsd:double.
    3. Else, if value is a JSON number with no non-zero fractional part, create an RDF Literal lit using the canonical representation for value with datatype xsd:integer.
    4. Otherwise, create an RDF Literal lit using the canonical representation for value with datatype xsd:string.

    Emit the following triple:

    subject
    node subject
    predicate
    property
    object
    literal node lit

Examples

In addition to the namespaces defined above, the examples provided here make use of the following namespaces:

dc:
http://purl.org/dc/terms/
foaf:
http://xmlns.com/foaf/0.1/
oa:
http://www.w3.org/ns/oa#
org:
http://www.w3.org/ns/org#
schema:
http://schema.org/

Furthermore, these examples also make use of the Turtle syntax @base declaration (as defined in [[turtle]]). Where a single tabular data file is used in the example, the @base declaration is set to the URL of that tabular data file.

Each of the examples expresses more complex conversions - it is recommended that readers of this specification work through the examples in sequential order.

Simple example

This example comprises a single annotated table containing information attributes about countries; country code, position (latitude, longitude) and name. Whilst the input tabular data file, published at http://example.org/countries.csv, includes a header line, no further metadata annotations are given. The tabular data file is provided below:

        

The annotated table generated from parsing the tabular data file is shown below and provides the basis for the conversion to RDF.

Annotations for the resulting table T, with 4 columns and 3 rows, are shown below:

idcore annotations
urlcolumnsrows
Thttp://example.org/countries.csvC1, C2, C3, C4R1, R2, R3

Annotations for the columns, rows and cells in table T are shown in the tables below.

Column annotations:

idcore annotations
tablenumbersource numbercellsnametitles
C1T11C1.1, C2.1, C3.1countryCodecountryCode
C2T22C1.2, C2.2, C3.2latitudelatitude
C3T33C1.3, C2.3, C3.3longitudelongitude
C4T44C1.4, C2.4, C3.4namename

Row annotations:

idcore annotations
tablenumbersource numbercells
R1T12C1.1, C1.2, C1.3, C1.4
R2T23C2.1, C2.2, C2.3, C2.4
R3T34C3.1, C3.2, C3.3, C3.4

Cell annotations:

idcore annotations
tablecolumnrowstring valuevalueproperty URL
C1.1TC1R1"AD""AD"null
C1.2TC2R1"42.5""42.5"null
C1.3TC3R1"1.6""1.6"null
C1.4TC4R1"Andorra""Andorra"null
C2.1TC1R2"AE""AE"null
C2.2TC2R2"23.4""23.4"null
C2.3TC3R2"53.8""53.8"null
C2.4TC4R2"United Arab Emirates""United Arab Emirates"null
C3.1TC1R3"AF""AF"null
C3.2TC2R3"33.9""33.9"null
C3.3TC3R3"67.7""67.7"null
C3.4TC4R3"Afghanistan""Afghanistan"null

Minimal mode output for this example is provided in Turtle [[turtle]] syntax below:

        

The about URL annotation has not been set for cells in table T ({ "url": "http://example.org/countries.csv"}); cells in a given row where about URL has not been specified are assumed to refer to the same subject. This unspecified subject is treated as a blank node.

Given that the property URL is null for cells in table T ({ "url": "http://example.org/countries.csv"}), the property URL defaults to the URI Template (see [[RFC6570]]) #{[column-name]}, where [column-name] is the value of the name annotation of the column associated with the cell. For example, the value of the property URL annotation for all cells in column C1 ("name": "countryCode") is http://example.org/countries.csv#countryCode.

Standard mode output for this example is provided in Turtle [[turtle]] syntax below:

        

Even though the table was defined in isolation, the annotated table is wrapped in a group of tables.

The type of both table and group of tables objects is explicitly stated; csvw:TableGroup and csvw:Table respectively.

The csvw:url property provides reference to the original tabular data file and to specific rows therein - noting the need to escape the Turtle-syntax reserved character = (U+003D) within the fragment identifier.

The row number is provided for each row using csvw:rownum property.

A subject and row are related using the csvw:describes property.

Example with single table and rich annotations

This example is based on Use Case #11 - City of Palo Alto Tree Data and comprises a single annotated table describing an inventory of tree maintenance operations. The input tabular data file, published at http://example.org/tree-ops-ext.csv, and the associated metadata description http://example.org/tree-ops-ext.csv-metadata.json are provided below:

        
        

The notes annotation in the metadata description uses the Open Annotation data model currently under development within the Web Annotations Working Group. This is purely illustrative; no constraints are placed on the value of the notes annotation.

The annotated table generated from parsing the tabular data file and associated metadata is shown below and provides the basis for the conversion to RDF.

Core annotations for the resulting table T, with 9 columns and 3 rows, are shown below:

idcore annotations
idurlcolumnsrowsnotes
T<http://example.org/tree-ops-ext>http://example.org/tree-ops-ext.csvC1, C2, C3, C4, C5, C6, C7, C8, C9R1, R2, R3[{ "@type": "oa:Annotation", ... }]

Non-core annotations for the table T are:

dc:title
"Tree Operations"
dcat:keyword
["tree", "street", "maintenance"]
dc:publisher
[{ "schema:name": "Example Municipality", "schema:url": { "@id": "http://example.org" } }]
dc:license
{ "@id": "http://opendefinition.org/licenses/cc-by/" }
dc:modified
"2010-12-31"

The value of the notes annotation has been shortened for clarity in the table above.

Annotations for the columns, rows and cells in table T are shown in the tables below.

Column annotations:

idcore annotationsannotations
tablenumbersource numbercellsnametitlesrequiredsuppress outputdc:description
C1T11C1.1, C2.1, C3.1GIDGID, Generic IdentifiertruetrueAn identifier for the operation on a tree.
C2T22C1.2, C2.2, C3.2on_streetOn StreetThe street that the tree is on.
C3T33C1.3, C2.3, C3.3speciesSpeciesThe species of the tree.
C4T44C1.4, C2.4, C3.4trim_cycleTrim CycleThe operation performed on the tree.
C5T55C1.5, C2.5, C3.5dbhDiameter at Breast HtDiameter at Breast Height (DBH) of the tree (in feet), measured 4.5ft above ground.
C6T66C1.6, C2.6, C3.6inventory_dateInventory DateThe date of the operation that was performed.
C7T77C1.7, C2.7, C3.7commentsCommentsSupplementary comments relating to the operation or tree.
C8T88C1.8, C2.8, C3.8protectedProtectedIndication (YES / NO) whether the tree is subject to a protection order.
C9T99C1.9, C2.9, C3.9kmlKMLKML-encoded description of tree location.

In this example, output for column C1 (GID) is not required; note the suppress output annotation on this column.

Row annotations:

idcore annotations
tablenumbersource numbercellsprimary key
R1T12C1.1, C1.2, C1.3, C1.4, C1.5, C1.6, C1.7, C1.8, C1.9C1.1
R2T23C2.1, C2.2, C2.3, C2.4, C2.5, C2.6, C2.7, C2.8, C2.9C2.1
R3T34C3.1, C3.2, C3.3, C3.4, C3.5, C3.6, C3.7, C3.8, C3.9C3.1

Cell annotations:

idcore annotations
tablecolumnrowstring valuevalueabout URL
C1.1TC1R1"1""1"http://example.org/tree-ops-ext#gid-1
C1.2TC2R1"ADDISON AV""ADDISON AV"<http://example.org/tree-ops-ext#gid-1>
C1.3TC3R1"Celtis australis""Celtis australis"<http://example.org/tree-ops-ext#gid-1>
C1.4TC4R1"Large Tree Routine Prune""Large Tree Routine Prune" (English)<http://example.org/tree-ops-ext#gid-1>
C1.5TC5R1"11"11<http://example.org/tree-ops-ext#gid-1>
C1.6TC6R1"10/18/2010"2010-10-18<http://example.org/tree-ops-ext#gid-1>
C1.7TC7R1""null<http://example.org/tree-ops-ext#gid-1>
C1.8TC8R1""false<http://example.org/tree-ops-ext#gid-1>
C1.9TC9R1"<Point><coordinates>-122.156485,37.440963</coordinates></Point>""<Point><coordinates>-122.156485,37.440963</coordinates></Point>" (XML)<http://example.org/tree-ops-ext#gid-1>
C2.1TC1R2"2""2"<http://example.org/tree-ops-ext#gid-2>
C2.2TC2R2"EMERSON ST""EMERSON ST"<http://example.org/tree-ops-ext#gid-2>
C2.3TC3R2"Liquidambar styraciflua""Liquidambar styraciflua"<http://example.org/tree-ops-ext#gid-2>
C2.4TC4R2"Large Tree Routine Prune""Large Tree Routine Prune" (English)<http://example.org/tree-ops-ext#gid-2>
C2.5TC5R2"11"11<http://example.org/tree-ops-ext#gid-2>
C2.6TC6R2"6/2/2010"2010-06-02<http://example.org/tree-ops-ext#gid-2>
C2.7TC7R2""null<http://example.org/tree-ops-ext#gid-2>
C2.8TC8R2""false<http://example.org/tree-ops-ext#gid-2>
C2.9TC9R2"<Point><coordinates>-122.156749,37.440958</coordinates></Point>""<Point><coordinates>-122.156749,37.440958</coordinates></Point>" (XML)<http://example.org/tree-ops-ext#gid-2>
C3.1TC1R3"6""6"<http://example.org/tree-ops-ext#gid-6>
C3.2TC2R3"ADDISON AV""ADDISON AV"<http://example.org/tree-ops-ext#gid-6>
C3.3TC3R3"Robinia pseudoacacia""Robinia pseudoacacia"<http://example.org/tree-ops-ext#gid-6>
C3.4TC4R3"Large Tree Routine Prune""Large Tree Routine Prune" (English)<http://example.org/tree-ops-ext#gid-6>
C3.5TC5R3"29"29<http://example.org/tree-ops-ext#gid-6>
C3.6TC6R3"6/1/2010"2010-06-01<http://example.org/tree-ops-ext#gid-6>
C3.7TC7R3"cavity or decay; trunk decay; codominant leaders; included bark; large leader or limb decay; previous failure root damage; root decay; beware of BEES""cavity or decay", "trunk decay", "codominant leaders", "included bark", "large leader or limb decay", "previous failure root damage", "root decay", "beware of BEES"<http://example.org/tree-ops-ext#gid-6>
C3.8TC8R3"YES"true<http://example.org/tree-ops-ext#gid-6>
C3.9TC9R3"<Point><coordinates>-122.156299,37.441151</coordinates></Point>""<Point><coordinates>-122.156299,37.441151</coordinates></Point>" (XML)<http://example.org/tree-ops-ext#gid-6>

The lists of values from cells in column C7 ("name": "comments") are assumed to be unordered as the boolean ordered annotation, which defaults to false, has not be set within the metadata description.

Minimal mode output for this example is provided in Turtle [[turtle]] syntax below:

        

The subject described by each row is explicitly defined using the about URL annotation; e.g. the subject of row R1 is http://example.org/tree-ops-ext#gid-1.

Output for column C1 ({ "name": "GID" }) is not included as the column suppress output annotation is true.

A language tag is specified for values of column C4 ({ "name": "trim_cycle" }) as the cell value language annotation is en.

The datatype annotation is set on columns C5, C6, C8 and C9 ({ "name": "dbh"}, { "name": "inventory_date" }, { "name": "protected" } and { "name": "kml" }); integer, date, boolean and xml respectively. The datatype property is inherited by all cells in each of those columns, therefore the RDF output for those cells includes the appropriate datatype IRI.

Cells C1.7 and C2.7 (rows R1 and R2; column, { "name": "comments" }) have null values - no output is included for these cells.

Cell C3.7 (row R3; column, { "name": "comments" }) contains an unordered sequence of values; the set of values are included as a simple set of triples as opposed to an instance of rdf:List as the ordered annotation has defaulted to false.

Standard mode output for this example is provided in Turtle [[turtle]] syntax below:

        

Table T ({ "url": "http://example.org/tree-ops-ext.csv"}) has been explicitly identified: { "@id": "<http://exmple.org/tree-ops-ext>"}.

Non-core annotations and notes specified for table T ({ "url": "http://example.org/tree-ops-ext.csv"}) are included in the output.

As the metadata description file http://example.org/tree-ops-ext.csv-metadata.json defines a default language within the context ("@context": ["http://www.w3.org/ns/csvw", {"@language": "en"}]), all non-core annotations of type string (e.g. dc:title, dcat:keyword, dc:publisher, dc:license and dc:modified) are expressed in the RDF output using the the appropriate language tag.

Example with single table and using virtual columns to produce multiple subjects per row

This example uses a single annotated table describing a listing of music events. Each row from the tabular data file corresponds to three resources; the music event itself, the location where that event occurs and the offer to sell tickets for that event. The goal is to convert the CSV content into schema.org markup that a search engine such as Google can use to index music events. Details of how Google expects this information to be structured can be found here.

The input tabular data file, published at http://example.org/events-listing.csv, and the associated metadata description http://example.org/events-listing.csv-metadata.json are provided below:

        
        

The CSV to RDF translation is limited to providing one statement, or triple, per column in the table. The target schema.org markup requires 10 statements to describe each event. As the base tabular data file contains 5 columns, an additional 5 virtual columns have been added in order to provide for the full complement of statements—including the relationships between the 3 resources (event, location, and offer) described by each row of the table. Note that the virtual annotation is set to true for these virtual columns.

Furthermore, note that no attempt is made to reconcile between locations or offers that may be associated with more than one event; every row in the table will create both a new location resource and offer resource in addition to the event resource. If considered necessary, applications such as OpenRefine may be used to identify and reconcile duplicate location resources once the RDF output has been generated.

The annotated table generated from parsing the tabular data file and associated metadata is shown below and provides the basis for the conversion to RDF.

Annotations for the resulting table T, with 10 columns and 2 rows, are shown below:

idcore annotations
urlcolumnsrows
Thttp://example.org/events-listing.csvC1, C2, C3, C4, C5, C6, C7, C8, C9, C10R1, R2

Annotations for the columns, rows and cells in table T are shown in the tables below.

Column annotations:

C1T11C1.1, C2.1nameName
C2T22C1.2, C2.2start_dateStart Date
C3T33C1.3, C2.3location_nameLocation Name
C4T44C1.4, C2.4location_addressLocation Address
C5T55C1.5, C2.5ticket_urlTicket Url
C6T66C1.6, C2.6type_eventtrue
C7T77C1.7, C2.7type_placetrue
C8T88C1.8, C2.8type_offertrue
C9T99C1.9, C2.9locationtrue
C10T1010C1.10, C2.10offerstrue

Row annotations:

idcore annotations
tablenumbersource numbercells
R1T12C1.1, C1.2, C1.3, C1.4, C1.5, C1.6, C1.7, C1.8, C1.9, C1.10
R2T23C2.1, C2.2, C2.3, C2.4, C2.5, C2.6, C2.7, C2.8, C2.9, C2.10

Cell annotations:

idcore annotations
tablecolumnrowstring valuevalueabout URLproperty URLvalue URL
C1.1TC1R1"B.B. King""B.B. King"<http://example.org/events-listing.csv#event-1>schema:name
C1.2TC2R1"2014-04-12T19:30"2014-04-12T19:30:00<http://example.org/events-listing.csv#event-1>schema:startDate
C1.3TC3R1"Lupo’s Heartbreak Hotel""Lupo’s Heartbreak Hotel"<http://example.org/events-listing.csv#place-1>schema:name
C1.4TC4R1"79 Washington St., Providence, RI""79 Washington St., Providence, RI"<http://example.org/events-listing.csv#place-1>schema:address
C1.5TC5R1"https://www.etix.com/ticket/1771656"<https://www.etix.com/ticket/1771656><http://example.org/events-listing.csv#offer-1>schema:url
C1.6TC6R1""null<http://example.org/events-listing.csv#event-1>rdf:typeschema:MusicEvent
C1.7TC7R1""null<http://example.org/events-listing.csv#place-1>rdf:typeschema:Place
C1.8TC8R1""null<http://example.org/events-listing.csv#offer-1>rdf:typeschema:Offer
C1.9TC9R1""null<http://example.org/events-listing.csv#event-1>schema:location<http://example.org/events-listing.csv#place-1>
C1.10TC10R1""null<http://example.org/events-listing.csv#event-1>schema:offers<http://example.org/events-listing.csv#offer-1>
C2.1TC1R2"B.B. King""B.B. King"<http://example.org/events-listing.csv#event-2>schema:name
C2.2TC2R2"2014-04-13T20:00"2014-04-13T20:00:00<http://example.org/events-listing.csv#event-2>schema:startDate
C2.3TC3R2"Lynn Auditorium""Lynn Auditorium"<http://example.org/events-listing.csv#place-2>schema:name
C2.4TC4R2"Lynn, MA, 01901""Lynn, MA, 01901"<http://example.org/events-listing.csv#place-2>schema:address
C2.5TC5R2"http://frontgatetickets.com/venue.php?id=11766"<http://frontgatetickets.com/venue.php?id=11766><http://example.org/events-listing.csv#offer-2>schema:url
C2.6TC6R2""null<http://example.org/events-listing.csv#event-2>rdf:typeschema:MusicEvent
C2.7TC7R2""null<http://example.org/events-listing.csv#place-2>rdf:typeschema:Place
C2.8TC8R2""null<http://example.org/events-listing.csv#offer-2>rdf:typeschema:Offer
C2.9TC9R2""null<http://example.org/events-listing.csv#event-2>schema:location<http://example.org/events-listing.csv#place-2>
C2.10TC10R2""null<http://example.org/events-listing.csv#event-2>schema:offers<http://example.org/events-listing.csv#offer-2>

Minimal mode output for this example is provided in Turtle [[turtle]] syntax below:

        

Three resources are defined for each row within the table; event, location, and offer.

Each column description in the metadata explicitly defines both aboutUrl and propertyUrl properties which are used to create the about URL and property URL annotations on the column's cells.

Columns C6, C7 and C8 ({ "name": "type_event"}, { "name": "type_place"} and { "name": "type_offer"}) define the semantic types of the resources described by each row: schema:MusicEvent, schema:Place and schema:Offer respectively.

Column C9 ({ "name": "location"}) uses the about URL, property URL and value URL to assert the relationship between the event and location resources.

Column C10 ({ "name": "offer"}) uses the about URL, property URL and value URL to assert the relationship between the event and offer resources.

Standard mode output for this example is provided in Turtle [[turtle]] syntax below:

        

The resources described by each row are explcitly defined using the about URL annotation this case three resources per row (event, location, and offer). The relationship between the row and each subject resource is asserted using the csvw:describes property; e.g. for row R1 we state [] csvw:describes t1:event-1, t1:place-1, t1:offer-1 . .

Example with group of tables comprising four interrelated tables

This example is based on Use Case #4 - Publication of public sector roles and salaries and uses four annotated tables published as a group of tables. Information about senior roles and junior roles within a government department or organization are published in CSV format by each department. These are validated against a centrally published schema to ensure that all the data published by departments is consistent. Additionally, lists of organizations and professions are also published centrally, providing controlled vocabularies against which departmental submissions are validated.

Information published about junior and senior roles provides summary information for each post within the government department or organization. Whilst the junior role information is anonymous, providing only an indication of the number of full-time-equivalent (FTE) staff occupying a given post, the senior role information specifies the named individual occupying each post. As such, each row from the tabular data file describing senior roles corresponds to two resources; the post and the person occupying that post.

This example is concerned only with converting the information provided each government department or organization not the centrally published information listing organizations and professions.

The input tabular data files and associated metadata descriptions are provided below:

        
        
        
        
        
        
        
        
        

This example makes extensive use of the example.org domain. As described in [[RFC6761]], this domain is used for illustrative examples within documentation. In reality, the resources described here with the URL path /gov.uk would be centrally published by the UK Government at, say, the domain data.gov.uk.

Given that these resources are centrally published with an aspiration for reuse, the schema descriptions have been factored out into separate resources. As such, the top-level metadata description resource metadata.json simply provides the list of tables and binds each of them to the appropriate schema that is defined elsewhere.

Finally, note that because the centrally published metadata descriptions are intended to be reused across many government departments and organizations, extra consideration has been given to defining URIs for the person and post resources defined in each row of the senior roles tabular data and subsequently referenced from the junior roles tabular data. To ensure that naming clashes are avoided, the unique reference for the organization to which the person or post belongs has been included in a path segment of the identifier. For example, the URI template property aboutUrl used to identify the senior post is specified as http://example.org/organization/{organizationRef}/post/{ref}, thus yielding the URI http://example.org/organization/hefce.ac.uk/post/90115 for the post described in the first row of the senior roles tabular data.

The group of tables generated from parsing the tabular data files and associated metadata is shown below and provides the basis for the conversion to RDF.

Annotations for the group of tables G and the four tables Ta, Tb, Tc, and Td are shown below.

Group of Tables annotations:

idcore annotations
tables
GTa, Tb, Tc, Td

Table annotations:

idcore annotations
urlcolumnsrowssuppress outputforeign keys
Tahttp://example.org/gov.uk/data/organizations.csvCa1, Ca2, Ca3Ra1, Ra2trueFa1
Tbhttp://example.org/gov.uk/professions.csvCb1Rb1, Rb2, Rb3, Rb4true
Tchttp://example.org/senior-roles.csvCc1, Cc2, Cc3, Cc4, Cc5, Cc6Rc1, Rc2falseFc1, Fc2, Fc3
Tdhttp://example.org/junior-roles.csvCd1, Cd2, Cd3, Cd4, Cd5, Cd6, Cd7Rd1, Rd2falseFd1, Fd2, Fd3

In this example, output for the centrally published lists of organizations and professions, tables Ta and Tb (http://example.org/gov.uk/data/organizations.csv and http://example.org/gov.uk/data/professions.csv respectively), are not required; only information from the departmental submissions is to be translated to RDF. Note the suppress output annotation on this table.

The following foreign keys are defined:

idcolumns in tablecolumns in referenced table
Fa1Ca3Ca1
Fc1Cc5Cc1
Fc2Cc6Cb1
Fc3Cc7Ca1
Fd1Cd1Cc1
Fd2Cd7Cb1
Fd3Cd8Ca1

Annotations for the columns, rows and cells in table T are shown in the tables below.

Column annotations:

idcore annotations
tablenumbersource numbercellsnametitlesrequiredvirtual
Ca1Ta11Ca1.1, Ca2.1refOrganization Unique Referencetrue
Ca2Ta11Ca1.2, Ca2.2nameOrganization Name
Ca3Ta11Ca1.3, Ca2.3departmentDepartment Reference
Cb1Tb11Cb1.1, Cb2.1, Cb3.1, Cb4.1nameProfessiontrue
Cc1Tc11Cc1.1, Cc2.1refPost Unique Referencetrue
Cc2Tc22Cc1.2, Cc2.2nameName
Cc3Tc33Cc1.3, Cc2.3gradeGrade
Cc4Tc44Cc1.4, Cc2.4jobJob Title
Cc5Tc55Cc1.5, Cc2.5reportsToReports to Senior Post
Cc6Tc66Cc1.6, Cc2.6professionProfession
Cc7Tc77Cc1.7, Cc2.7organizationRefOrganization Referencetrue
Cc8Tc88Cc1.8, Cc2.8post_holdertrue
Cd1Td11Cd1.1, Cd2.1reportsToSeniorReporting Senior Posttrue
Cd2Td22Cd1.2, Cd2.2gradeGrade
Cd3Td33Cd1.3, Cd2.3min_payPayscale Minimum (£)
Cd4Td44Cd1.4, Cd2.4max_payPayscale Maximum (£)
Cd5Td55Cd1.5, Cd2.5jobGeneric Job Title
Cd6Td66Cd1.6, Cd2.6numberNumber of Posts (FTE)
Cd7Td77Cd1.7, Cd2.7professionProfession
Cd8Td88Cd1.8, Cd2.8organizationRefOrganization Referencetrue

Column Cc8, with the virtual annotation specified as true, is used to relate the person resource, whose name is provided in column Cc2, to the associated post resource within the current row of table Tc ({ "url": "http://example.org/senior-roles.csv" }).

Row annotations:

idcore annotations
tablenumbersource numbercells
Ra1Ta12Ca1.1, Ca1.2, Ca1.3
Ra2Ta23Ca2.1, Ca2.2, Ca2.3
Rb1Tb12Cb1.1
Rb2Tb23Cb2.1
Rb3Tb34Cb3.1
Rb4Tb45Cb4.1
Rc1Tc12Cc1.1, Cc1.2, Cc1.3, Cc1.4, Cc1.5, Cc1.6, Cc1.7, Cc1.8
Rc2Tc23Cc2.1, Cc2.2, Cc2.3, Cc2.4, Cc2.5, Cc2.6, Cc2.7, Cc2.8
Rd1Td12Cd1.1, Cd1.2, Cd1.3, Cd1.4, Cd1.5, Cd1.6, Cd1.7, Cd1.8
Rd2Td23Cd2.1, Cd2.2, Cd2.3, Cd2.4, Cd2.5, Cd2.6, Cd2.7, Cd2.8

Cell annotations:

idcore annotations
tablecolumnrowstring valuevalueabout URLproperty URLvalue URL
Ca1.1TaCa1Ra1"hefce.ac.uk""hefce.ac.uk"<http://example.org/organization/hefce.ac.uk>dc:identifier
Ca1.2TaCa2Ra1"Higher Education Funding Council for England""Higher Education Funding Council for England"<http://example.org/organization/hefce.ac.uk>foaf:name
Ca1.3TaCa3Ra1"bis.gov.uk""bis.gov.uk"<http://example.org/organization/hefce.ac.uk>org:subOrganizationOf<http://example.org/organization/bis.gov.uk>
Ca2.1TaCa1Ra2"bis.gov.uk""bis.gov.uk"<http://example.org/organization/bis.gov.uk>dc:identifier
Ca2.2TaCa2Ra2"Department for Business, Innovation and Skills""Department for Business, Innovation and Skills"<http://example.org/organization/bis.gov.uk>foaf:name
Ca2.3TaCa3Ra2"xx"null<http://example.org/organization/bis.gov.uk>org:subOrganizationOf
Cb1.1TbCb1Rb1"Finance""Finance"
Cb2.1TbCb1Rb2"Information Technology""Information Techology"
Cb3.1TbCb1Rb3"Operational Delivery""Operational Delivery"
Cb4.1TbCb1Rb4"Policy""Policy"
Cc1.1TcCc1Rc1"90115""90115"<http://example.org/organization/hefce.ac.uk/post/90115>dc:identifier
Cc1.2TcCc2Rc1"Steve Egan""Steve Egan"<http://example.org/organization/hefce.ac.uk/person/1>foaf:name
Cc1.3TcCc3Rc1"SCS1A""SCS1A"<http://example.org/organization/hefce.ac.uk/post/90115><http://example.org/gov.uk/def/grade>
Cc1.4TcCc4Rc1"Deputy Chief Executive""Deputy Chief Executive"<http://example.org/organization/hefce.ac.uk/post/90115><http://example.org/gov.uk/def/job>
Cc1.5TcCc5Rc1"90334""90334"<http://example.org/organization/hefce.ac.uk/post/90115>org:reportsTo<http://example.org/organization/hefce.ac.uk/post/90334>
Cc1.6TcCc6Rc1"Finance""Finance"<http://example.org/organization/hefce.ac.uk/post/90115><http://example.org/gov.uk/def/profession>
Cc1.7TcCc7Rc1"hefce.ac.uk""hefce.ac.uk"<http://example.org/organization/hefce.ac.uk/post/90115>org:postIn<http://example.org/organization/hefce.ac.uk>
Cc1.8TcCc8Rc1""null<http://example.org/organization/hefce.ac.uk/post/90115>org:heldBy<http://example.org/organization/hefce.ac.uk/person/1>
Cc2.1TcCc1Rc2"90334""90334"<http://example.org/organization/hefce.ac.uk/post/90334>dc:identifier
Cc2.2TcCc2Rc2"Sir Alan Langlands""Sir Alan Langlands"<http://example.org/organization/hefce.ac.uk/person/2>foaf:name
Cc2.3TcCc3Rc2"SCS4""SCS4"<http://example.org/organization/hefce.ac.uk/post/90334><http://example.org/gov.uk/def/grade>
Cc2.4TcCc4Rc2"Chief Executive""Chief Executive"<http://example.org/organization/hefce.ac.uk/post/90334><http://example.org/gov.uk/def/job>
Cc2.5TcCc5Rc2"xx"null<http://example.org/organization/hefce.ac.uk/post/90334>org:reportsTo
Cc2.6TcCc6Rc2"Policy""Policy"<http://example.org/organization/hefce.ac.uk/post/90334><http://example.org/gov.uk/def/profession>
Cc2.7TcCc7Rc2"hefce.ac.uk""hefce.ac.uk"<http://example.org/organization/hefce.ac.uk/post/90334>org:postIn<http://example.org/organization/hefce.ac.uk>
Cc2.8TcCc8Rc2""null<http://example.org/organization/hefce.ac.uk/post/90334>org:heldBy<http://example.org/organization/hefce.ac.uk/person/2>
Cd1.1TdCd1Rd1"90115""90115"org:reportsTo<http://example.org/organization/hefce.ac.uk/post/90115>
Cd1.2TdCd2Rd1"4""4"<http://example.org/gov.uk/def/grade>
Cd1.3TdCd3Rd1"17426"17426<http://example.org/gov.uk/def/min_pay>
Cd1.4TdCd4Rd1"20002"20002<http://example.org/gov.uk/def/max_pay>
Cd1.5TdCd5Rd1"Administrator""Administrator"<http://example.org/gov.uk/def/job>
Cd1.6TdCd6Rd1"8.67"8.67<http://example.org/gov.uk/def/number_of_posts>
Cd1.7TdCd7Rd1"Operational Delivery""Operational Delivery"<http://example.org/gov.uk/def/profession>
Cd1.8TdCd8Rd1"hefce.ac.uk""hefce.ac.uk"org:postIn<http://example.org/organization/hefce.ac.uk>
Cd2.1TdCd1Rd2"90115""90115"org:reportsTo<http://example.org/organization/hefce.ac.uk/post/90115>
Cd2.2TdCd2Rd2"5""5"<http://example.org/gov.uk/def/grade>
Cd2.3TdCd3Rd2"19546"19546<http://example.org/gov.uk/def/min_pay>
Cd2.4TdCd4Rd2"22478"22478<http://example.org/gov.uk/def/max_pay>
Cd2.5TdCd5Rd2"Administrator""Administrator"<http://example.org/gov.uk/def/job>
Cd2.6TdCd6Rd2"0.5"0.5<http://example.org/gov.uk/def/number_of_posts>
Cd2.7TdCd7Rd2"Operational Delivery""Operational Delivery"<http://example.org/gov.uk/def/profession>
Cd2.8TdCd8Rd2"hefce.ac.uk""hefce.ac.uk"org:postIn<http://example.org/organization/hefce.ac.uk>

Notice that value URL is not specified for cells Ca2.3 and Cc2.5 because in each case the cell value is null and the virtual annotation of column Cb5 is not defined.

Minimal mode output for this example is provided in Turtle [[turtle]] syntax below:

        

Output for tables Ta and Tb ({ "url": "http://example.org/gov.uk/data/organizations.csv" } and { "url": "http://example.org/gov.uk/data/professions.csv" }) are not included as the suppress output annotation is true.

The property URL is specified for all cells in tables Tc and Td.

Columns Cc5 and Cd1 ({ "name": "reportsTo" } and { "name": "reportsToSenior" }) use the about URL, property URL and value URL annotations to assert the relationship between the given post and the senior post it reports to for the cells therein.

Similarly, columns Cc7 and Cd8 (both with { "name": "organizationRef" }) use the about URL, property URL and value URL annotations to assert the relationship between the given post and the organization to which it belongs for the cells those columns.

Finally, note that two resources are created for each row within table Tc ({ "url": "http://example.org/senior-roles.csv" }): the person and the post they occupy. The relationship between these resources is specified via virtual column Cc8 ({ "name": "post_holder" }) using the about URL, property URL and value URL annotations.

Standard mode output for this example is provided in Turtle [[turtle]] syntax below:

        

Table group G was explicitly defined, but has not been explicitly identified; the table group and table resources are treated as blank nodes.

The person and post resources described by each row of table Tc ({ "url": "http://example.org/senior-roles.csv"}) are explcitly defined using the aboutUrl property; therefore, say, for row Rc1 we state [] csvw:describes <http://example.org/organization/hefce.ac.uk/post/90115>, <http://example.org/organization/hefce.ac.uk/person/1> .; whilst the aboutUrl property has not been defined for resources described by each row of table Td ({ "url": "http://example.org/junior-roles.csv"}); therefore blank nodes are used, e.g. for row Rd1 we state [] csvw:describes _:d8b8e40c-8c74-458b-99f7-64d1cf5c65f2 . .

Relationship to RDB Direct Mapping

The "Direct Mapping of Relational Data to RDF" W3C Recommendation [[rdb-direct-mapping]] defines a simple transformation (referred to as direct mapping) from a relational representation of data to RDF. The direct mapping takes as input a relational database (data and schema), and generates an RDF graph called the direct graph. Tables in a relational databases bear a strong resemblence to tabular data as defined in [[tabular-data-model]]; this section highlights the similarities and differences between the direct mapping and the tabular data mapping defined by this document. The following statements summarize the relationships:

The rest of this section will provide some examples. These are somewhat simplified forms of the examples as used in the [[rdb-direct-mapping]] document.

Single table case

Consider a simple table in a relational database, called People:

PK
ID fname addr
7 Bob 18

A relational table always has a schema that the direct mapping makes use of. In this example, the schema defines that:

Using the schema information, the direct graph is as follows (see [[rdb-direct-mapping]] for further details):

<http://foo.example.org/DB/People/ID=7> rdf:type <http://foo.example.org/DB/People>;
  <http://foo.example.org/DB/People/#ID> 7;
  <http://foo.example.org/DB/People/#fname> "Bob";
  <http://foo.example.org/DB/People/#addr> 18.

where http://foo.example.org/DB/ is the URL for the database containing the People table.

When exporting the table into CSV the column names would naturally be mapped on the titles of the respective columns. The (minimal) tabular data mapping yields:

[
  <http://foo.example.org/CSV/People/#ID> "7";
  <http://foo.example.org/CSV/People/#fname> "Bob";
  <http://foo.example.org/CSV/People/#addr> "18";
]

where http://foo.example.org/CSV/People is the URL of the CSV file exported from the relational database.

Comparing the two conversion results:

It is, however, possible to add annotation to the tabular data so that the two graphs would be semantically equivalent. Indeed, consider the following metadata using the definitions of the [[tabular-metadata]] specification:

{
  "@context": "http://www.w3.org/ns/csvw",
  "tableSchema": {
     "url" : "http://foo.example.org/CSV/People",
     "aboutUrl" : "http://foo.example.org/CSV/People/ID={ID}",
     "columns": [{
        "name": "ID",
        "datatype" : "integer"
      }, {
        "name": "fname",
      }, {
        "name": "addr",
        "datatype" : "integer"
      }, {
        "name": "type",
        "virtual": true,
        "propertyUrl": "rdf:type",
        "valueUrl" : "http://foo.example.org/CSV/People"
    }],
  }
}

The metadata adds annotations on datatypes, adds the RDF typing triple explicitly (using a virtual column), and uses the aboutUrl URL template property to provide the common subject. Essentially, the metadata provides the information that the direct mapping retrieves from the table schema.

The [[tabular-metadata]] specification includes an annotation for primary keys. However, that annotation is used for validation and does not influence the tabular data mapping, hence it is not included in the example metadata.

A processor exporting the table into a CSV file could also generate such a metadata automatically.

Simple, two table case

Consider a two interrelated simple tables in a relational database, called People and Addresses, respectively:

PK → Addresses(ID)
ID fname addr
7 Bob 18
PK
ID city state
18 Cambridge MA

Beyond what is already specified for the People table the corresponding relational schema also specifies that:

Using the schema information, the direct graph is as follows (see [[rdb-direct-mapping]] for further details):

<http://foo.example.org/DB/People/ID=7> rdf:type <http://foo.example.org/DB/People>;
  <http://foo.example.org/DB/People#ID> 7;
  <http://foo.example.org/DB/People#fname> "Bob";
  <http://foo.example.org/DB/People#addr> 18;
  <http://foo.example.org/DB/People#ref-addr> <http://foo.example.org/DB/Addresses/ID=18>.

<http://foo.example.org/DB/Addresses/ID=18> rdf:type <http://foo.example.org/DB/Addresses>;
  <http://foo.example.org/DB/Addresses#ID> 18;
  <http://foo.example.org/DB/Addresses#city> "Cambridge";
  <http://foo.example.org/DB/Addresses#state> "MA".

Note the highlighted RDF triple linking to the relevant row of the Addresses table; this corresponds to the foreign key in the schema.

Using the following metadata the tabular data mapping yields a semantically equivalent result:

{
    "@context": "http://www.w3.org/ns/csvw",
    "resources" : [{
      "url": "http://foo.example.org/CSV/People",
      "aboutUrl" : "http://foo.example.org/CSV/People/ID={ID}",
      "tableSchema": {
        "columns": [{
          "name": "ID",
          "datatype": "integer"
        }, {
          "name": "fname",
        }, {
          "name": "addr",
          "datatype": "integer"
        }, {
          "name": "ref",
          "virtual": true,
          "propertyUrl": "http://foo.example.org/CSV/People#ref-addr",
          "valueUrl" : "http://foo.example.org/CSV/Addresses/ID={addr}"
        }, {
          "name": "type",
          "virtual": true,
          "propertyUrl": "rdf:type",
          "valueUrl" : "http://foo.example.org/CSV/People"
        }],
      }
    }, {
      "url": "http://foo.example.org/CSV/Addresses",
      "aboutUrl" : "http://foo.example.org/CSV/Addresses/ID={ID}",
      "tableSchema": [{
        "columns": [{
          "name": "ID",
          "datatype": "integer"
        }, {
          "name": "city",
        }, {
          "name": "state",
        }, {
          "name": "type",
          "virtual": true,
          "propertyUrl": "rdf:type",
          "valueUrl" : "http://foo.example.org/CSV/Addresses"
        }],
    }]
}

More complex example with three tables

Consider a case with three, interrelated simple tables in a relational database, called People, Addresses, and Departments, respectively:

PK → Addresses(ID) → Department(name,city)
ID fname addr deptName deptCity
7 Bob 18 accounting Cambridge
PK
ID city state
18 Cambridge MA
PK Unique Key
ID name city
23 accounting Cambridge

The corresponding relational schema specifies that

Using the schema information, the direct graph is as follows (see [[rdb-direct-mapping]] for further details):

<http://foo.example.org/DB/People/ID=7> rdf:type <http://foo.example.org/DB/People>;
  <http://foo.example.org/DB/People/#ID> 7;
  <http://foo.example.org/DB/People/#fname> "Bob";
  <http://foo.example.org/DB/People/#addr> 18;
  <http://foo.example.org/DB/People/#ref-addr> <http://foo.example.org/DB/Addresses/ID=18>;
  <http://foo.example.org/DB/People/#deptName> "accounting";
  <http://foo.example.org/DB/People/#deptCity> "Cambridge";
  <http://foo.example.org/DB/People/#ref-deptName;deptCity> <http://foo.example.org/DB/Department/ID=23>.

<http://foo.example.org/DB/Addresses/ID=18> rdf:type <http://foo.example.org/DB/Addresses>;
  <http://foo.example.org/DB/Addresses/#ID> 18;
  <http://foo.example.org/DB/Addresses/#city> "Cambridge";
  <http://foo.example.org/DB/Addresses/#addr> "MA".

<http://foo.example.org/DB/Departments/ID=23> rdf:type <http://foo.example.org/DB/Addresses>;
  <http://foo.example.org/DB/Departments/#ID> 23;
  <http://foo.example.org/DB/Departments/#name> "accounting";
  <http://foo.example.org/DB/Departments/#city> "Cambridge".

The major difference, compared to the simpler example with foreign keys is the usage of unique keys. To generate the right object URI in the highlighted statement of the direct graph the processor has to find that unique key combination in the Departments table and, using that combination, has to establish the subject for that specific row that can is to be used as an object URI in the highlighted statement. This can be done because the direct mapping processor has simultaneous access to several tables within the same relational database. It is therefore straightforward to access all the tables in parallel and establish the necessary relationships to generate the triples.

However, this combination cannot be handled by the tabular data mapping. The situation for tabular data is indeed different: tables are typically generated from single and, potentially, very large CSV files, meaning that a tabular data mapping processor cannot be expected to handle several tables in parallel. That is the reason why the [[tabular-data-model]] does not include features that would require such parallel access. However, as a consequence, the output of the direct mapping for such tables cannot be reproduced by the tabular data mapping.

Note that the [[tabular-data-model]] includes an annotation for transformations; implementations may include scripts or templates to transform the output of the tabular data mapping and generate the required RDF graph.

Acknowledgements

Changes since previous versions

Changes since the working draft of 16 April 2015

Changes since the first public working draft of 08 January 2015

The document has undergone substantial changes since the first public working draft. Below are some of the changes made: