- Part One: About xAPI Profiles
- Part Two: xAPI Profiles Document Structure Specification
- 1.0. Reference Specifications
- 2.0. Technical Foundations
- 3.0. Structure
- 4.0. Document Interpretation and General Restrictions
- 5.0. Using Profiles in Statements
- 6.0. Profile Properties
- 7.0. Concepts
- 7.1. Verbs, Activity Types, and Attachment Usage Types
- 7.2. Extensions
- 7.3. Document Resources
- 7.4. Activities
- 8.0. Statement Templates
- 9.0. Patterns
- 10.0. The Context
- Part Three: xAPI Profiles Communication and Processing Specification
- 1.0. Profile Server
- 1.1. Profile Versions
- 1.2. Best Practices
- 1.3. Example SPARQL Queries
- 2.0. Algorithms
- 3.0. Libraries
- 1.0. Profile Server
- http://json-ld.org
- https://www.w3.org/TR/skos-reference/
- https://www.w3.org/TR/2013/REC-prov-dm-20130430/
- https://www.w3.org/TR/rdf-sparql-query/
This specification describes how to author an xAPI Profile. It describes a set of rules for authoring JSON, specifically JSON-LD. Since JSON-LD is a syntax for RDF, the resulting Profile is really a set of triples—subject, predicate, object—creating a semantic data set. However, for authoring, all that matters is following the rules given for JSON, which will lead to the richer semantic data naturally. Because of this, no JSON-LD processing is required by systems consuming xAPI Profiles, though there will be advantages to doing so for some purposes.
When a Profile is serialized into JSON, it MUST be consistent with what would be produced by the standard JSON-LD 1.1 Processing Algorithms and API Compaction Algorithm. The compaction must use, at least, the normative JSON-LD contexts provided with this specification. Following all the rules given in this document is sufficient to ensure that.
Under the hood, xAPI Profiles will use several well-established semantic web technologies: SKOS, to connect xAPI Concepts together, and PROV, to describe the provenance (most notably the versioning) of Profiles. Several properties in xAPI Profiles use names of properties from SKOS and PROV.
Profiles serve two primary technical goals. First, they contain metadata about xAPI Concepts intended for reuse within Statements, such as verbs and activity types. The metadata includes connections between Concepts, not just within the current Profile, but also as used in other Profiles, supporting a rich ecosystem of related terms. An xAPI Concept is any building block for use in Statements, and new versions of the Profile specification may introduce new Concepts that can be described. The basis for xAPI Concepts is the SKOS Concept, a flexible way to refer to "specific ideas or meanings established within a knowledge organization system."
Second, they contain specific rules about using those Concepts properly in specific situations, both on how to form individual Statements containing specific Concepts, and how to group those Statements together in Patterns of multiple Statements. These rules allow Profile authors to require specific elements, describe precise orderings, and many other options.
To assist in accomplishing these two primary goals, Profiles also contain metadata about themselves—descriptions, authorship, versioning, and so forth.
- All properties in tables are required in all cases unless marked optional.
- Properties marked optional may be required in some situations. If no additional information is provided on the usage of an optional property, including it or not is entirely up to the Profile author.
- All properties that are not JSON-LD keywords (or aliases thereof) MUST expand to absolute IRIs during processing as defined in the JSON-LD specification.
- All properties that are not JSON-LD keywords (or aliases thereof) and not described by this specification MUST be expressed using compact IRIs or absolute IRIs.
- JSON-LD keywords (or aliases thereof) that are not specified as properties in this document MAY be included anywhere they are legal in JSON-LD.
- Values in a Profile MUST NOT be: empty objects, null, empty strings, or empty arrays.
- All requirements on the structure of Profiles MUST be followed by Profile Authors.
- All requirements on Statements following Profiles MUST be followed by Learning Record Providers when authoring Statements and by Profile Validators when validating Statements.
Using an introduced Concept, such as an activity type, verb, attachment usage type, extension, activity, or document resource, can be done freely, provided the defined usage and meaning are adhered to. But a Learning Record Provider can go further, and make sure to adhere to Profile-described Statement Templates and Patterns. Learning Record Providers authoring Statements that conform to matching Profile-described Statement Templates and Patterns SHOULD include the most up-to-date conformant Profile version as a category context activity with id equal to the version's id
in those Statements, and Statements containing a Profile version as a category context activity MUST conform to any matching Statement Templates and Patterns that Profile version describes.
A Profile includes a variety of metadata, both natural language text for humans to understand the Profile, and structured data about versions and who created it. In addition to the metadata, there are properties for describing the Concepts, Statement Templates, and Patterns of the Profile.
Property | Type | Description | Required |
---|---|---|---|
id |
IRI | The IRI of the Profile overall (not a specific version) | Required |
@context |
URI | SHOULD be https://w3id.org/xapi/profiles/context and MUST contain this URI if array-valued. |
Required |
type |
String | MUST be Profile . |
Required |
conformsTo |
URI | Canonical URI of the Profile specification version conformed to. The Profile specification version of this document is https://w3id.org/xapi/profiles#1.0. | Required |
prefLabel |
Object | Language map of names for this Profile. | Required |
definition |
Object | Language map of descriptions for this Profile. If there are additional rules for the Profile as a whole that cannot be expressed using this specification, include them here, or at the seeAlso URL. | Required |
seeAlso |
URL | A URL containing information about the Profile. Recommended instead of especially long definitions. | Optional |
versions |
Array | An array of all Profile version objects for this Profile. | Required |
author |
Object | An Organization or Person. | Required |
concepts |
Array | An array of Concepts that make up this Profile. | Optional |
templates |
Array | An array of Statement Templates for this Profile. | Optional |
patterns |
Array | An array of Patterns for this Profile. | Optional |
When seeAlso
is provided definition
SHOULD only include a short description of the Profile to aid in discovery and display.
Profile version objects make it convenient to track version history for Profiles, following recommendations for SKOS concept schemes and PROV version tracking generally. By using versions this way, it is possible to answer precise questions such as “what version of this Profile was current on the 3rd of January last year?”. Lack of robust versioning is frequently identified as an issue with RDF data.
Property | Type | Description | Required |
---|---|---|---|
id |
IRI | The IRI of the version ID | Required |
wasRevisionOf |
Array | An array, usually of length one, of IRIs of all Profile versions this version was written as a revision of | Optional |
generatedAtTime |
Timestamp | The date this version was created on | Required |
wasRevisionOf
MUST be used with all versions that succeed other Profile versions.
wasRevisionOf
may sometimes contain multiple Profile versions to support the scenario where a Profile subsumes another. In this case, a new Profile version would be defined with the two (or more) contributing Profiles listed within the wasRevisionOf
array.
A Profile that's been revised twice will have an array of Profile versions that looks something like:
[
{
"id": "http://example.com/profiles/superheroes/v3",
"wasRevisionOf": ["http://example.com/profiles/superheroes/v2"],
"generatedAtTime": "2020-02-20T20:20:20Z"
},
{
"id": "http://example.com/profiles/superheroes/v2",
"wasRevisionOf": ["http://example.com/profiles/superheroes/v1"],
"generatedAtTime": "2010-01-15T03:14:15Z"
},
{
"id": "http://example.com/profiles/superheroes/v1",
"generatedAtTime": "2010-01-14T12:13:14Z"
}
]
Note: there is nothing special about the URI structure, such as the use of "v#" on the end. Any URIs are legal, so long as every Profile version id
is unique and different from the overall Profile URI. Using a predictable URI structure is a good idea, though.
Use one of these in the author
property to indicate the author of this Profile version.
Property | Type | Description | Required |
---|---|---|---|
type |
String | Organization or Person |
Required |
name |
String | A string with the name of the organization or person | Required |
url |
URL | A URL for the Person or Group. | Optional |
Concepts are building blocks for use and reuse in xAPI data and other Profiles. In the case of Verbs, Activity Types, Attachment Usage Types, and Activities, the Concept is "the thing", and when you use that Concept in xAPI you're using it directly. In the case of Document Resources and Extensions, the Concept is "the shape of the thing" that the identifier can be used to point at, and will be used with many different values xAPI data.
A Profile MUST NOT define a Concept that is defined in another Profile unless it supersedes all versions of the other Profile containing the Concept and indicates that in with wasRevisionOf
.
All Concepts in a Profile MUST follow the rules of one of the subsections within this section. Since the types listed in each subsection are exclusive and required, that will always distinguish which section applies.
Verb, Activity Type, and Attachment Usage Type Concepts share the same properties. They're all Concepts that make sense to relate semantically to others of the same type, such as indicating one is a narrower form of another.
Property | Type | Description | Required |
---|---|---|---|
id |
IRI | The IRI of this Concept | Required |
type |
String | Verb , ActivityType , or AttachmentUsageType |
Required |
inScheme |
IRI | The IRI of the specific Profile version currently being described | Required |
prefLabel |
Object | A Language Map of the preferred names in each language | Required |
definition |
Object | A Language Map of the precise definition, including how to use the Concept properly in Statements | Required |
deprecated |
Boolean | A boolean. If true, this Concept is deprecated. | Optional |
broader |
Array | An array of IRIs of Concepts of the same type from this Profile version that have a broader meaning. |
Optional |
broadMatch |
Array | An array of IRIs of Concepts of the same type from a different Profile that have a broader meaning. |
Optional |
narrower |
Array | An array of IRIs of Concepts of the same type from this Profile version that have a narrower meaning. |
Optional |
narrowMatch |
Array | An array of IRIs of Concepts of the same type from different Profiles that have narrower meanings. |
Optional |
related |
Array | An array of IRIs of Concepts of the same type from this Profile version that are close to this Concept's meaning. |
Optional |
relatedMatch |
Array | An array of IRIs of Concepts of the same type from a different Profile or a different version of the same Profile that has a related meaning that is not clearly narrower or broader. Useful to establish conceptual links between Profiles that can be used for discovery. |
Optional |
exactMatch |
Array | An array of IRIs of Concepts of the same type from a different Profile or a different version of the same Profile that have exactly the same meaning. |
Optional |
related
MUST only be used on Concepts that are deprecated to indicate possible replacement Concepts in the same Profile, if there are any.relatedMatch
SHOULD be used to connect possible replacement Concepts to removed Concepts from previous versions of the same Profile, and for possible replacement Concepts in other Profiles of deprecated Concepts, as well as other loose relations.exactMatch
SHOULD be used rarely, mostly to describe connections to vocabularies that are no longer managed and do not use good URLs.
A Profile for competitive events might have verbs in it such as:
{
"id": "http://example.org/profiles/sports/verbs/placed",
"type": "Verb",
"inScheme": "http://example.org/profiles/sports/v2",
"prefLabel": {
"en": "placed"
},
"definition": {
"en": "Indicates a person finished the event in a ranked order. \
Use with the 'placement' extension."
},
"broadMatch": ["http://adlnet.gov/expapi/verbs/completed"]
}
The use of broadMatch
means that the ADL's completed
verb is a more general term than placed
.
{
"id": "http://example.org/profiles/sports/verbs/medaled",
"type": "Verb",
"inScheme": "http://example.org/profiles/sports/v2",
"prefLabel": {
"en": "medaled"
},
"definition": {
"en": "Indicates a person received a medal in a particular event. \
Use with the 'placement' extension."
},
"broader": ["http://example.org/profiles/sports/verbs/placed"]
}
broader
here works like broadMatch
did above, but is used instead because the two verbs are in the same Profile.
{
"id": "http://example.org/profiles/sports/verbs/qualified",
"type": "Verb",
"inScheme": "http://example.org/profiles/sports/v2",
"prefLabel": {
"en": "qualified"
},
"definition": {
"en": "Indicates a person is eligible for the event in the object."
},
"relatedMatch": ["https://w3id.org/xapi/adl/verbs/satisfied"]
}
Qualifying for an event isn't the same as (or broader or narrower than) satisfying it's requirements in the way meant in the ADL profile, but they're related concepts. By creating relatedMatch
links a Profile makes it easier for people using Profiles to discover appropriate terms.
Property | Type | Description | Required |
---|---|---|---|
id |
IRI | The IRI of the extension, used as the extension key in xAPI | Required |
type |
String | ContextExtension , ResultExtension , or ActivityExtension |
Required |
inScheme |
IRI | The IRI of the specific Profile version currently being described | Required |
prefLabel |
Object | A Language Map of descriptive names for the extension | Required |
definition |
Object | A Language Map of descriptions of the purpose and usage of the extension | Required |
deprecated |
Boolean | A boolean. If true, this Concept is deprecated. | Optional |
recommendedActivityTypes |
Array | Only allowed on an ActivityExtension . An array of activity type URIs that this extension is recommended for use with (extending to narrower of the same). |
Optional |
recommendedVerbs |
Array | Only allowed on a ContextExtension or a ResultExtension . An array of verb URIs that this extension is recommended for use with (extending to narrower of the same). |
Optional |
context |
IRI | the IRI of a JSON-LD context for this extension | Optional |
schema |
IRI | the IRI for accessing a JSON Schema for this extension. The JSON Schema can be used to constrain the extension to a single type. | Optional |
inlineSchema |
Object | An alternate way to include a JSON Schema, as a string. | Optional |
Profiles MUST use at most one of schema
and inlineSchema
for Extensions. JSON Schema expressions SHOULD adhere to the Draft-07 version of the JSON Schema specification.
Statements including extensions defined in a Profile MUST:
- only use a ContextExtension in context
- only use a ResultExtension in result
- only use an ActivityExtension in an Activity Definition.
A Profile for competitive events might define an extension to represent placing, such as
{
"id": "http://example.org/profiles/sports/extensions/place",
"type": "ResultExtension",
"inScheme": "http://example.org/profiles/sports/v2",
"prefLabel": {
"en": "placement"
},
"definition": {
"en": "Defines the place a person received in an event. \
The value is an object with a required numerical rank \
and an optional medal as a string"
},
"recommendedVerbs": [
"http://example.org/profiles/sports/verbs/placed"
],
"inlineSchema": "{ \"type\": \"object\", \"properties\":{ \
\"rank\": {\"type\": \"number\", \"required\": true}, \
\"medal\": {\"type\": \"string\"}}}"
}
This extension includes a JSON Schema that systems handling Statements with it can use to validate the structure of extension values. Also, it recommends a verb to use it with. While it only mentions the placed
verb, if the medaled
verb is defined as given above, it is narrower than placed
and is recommended as well.
Property | Type | Description | Required |
---|---|---|---|
id |
IRI | The IRI of the document resource, used as the stateId/profileId in xAPI | Required |
type |
String | One of: StateResource , AgentProfileResource , ActivityProfileResource |
Required |
inScheme |
IRI | The IRI of the specific Profile version currently being described | Required |
prefLabel |
Object | A Language Map of descriptive names for the document resource | Required |
definition |
Object | A Language Map of descriptions of the purpose and usage of the document resource | Required |
contentType |
String | The media type for the resource, as described in RFC 2046 (e.g. application/json ). |
Required |
deprecated |
Boolean | A boolean. If true, this Concept is deprecated. | Optional |
context |
IRI | The IRI of a JSON-LD context for this document resource. | Optional |
schema |
IRI | the IRI for accessing a JSON Schema for this document resource. | Optional |
inlineSchema |
String | An alternate way to include a JSON Schema, as a string. | Optional |
Profiles MUST use at most one of schema
and inlineSchema
for Document Resources. JSON Schema expressions SHOULD adhere to the Draft-07 version of the JSON Schema specification.
Learning Record Store Clients sending Document Resources
- MUST use the
id
as thestateId
orprofileId
(as appropriate) when interacting with the corresponding resource. - MUST use the
contentType
given in the Content-Type header, including any parameters as given. - MAY add additional parameters to the Content-Type header that are not specified in the Concept.
- MUST
- only send a StateResource to a State Resource location
- only send an AgentProfileResource to an Agent Profile Resource location
- only send an ActivityProfileResource to an Activity Profile Resource location
Profile Validators receiving Document Resources MUST validate Learning Record Store Clients follow the requirements for Document Resources.
A Profile for competitive events might include a way to store preferred t-shirt sizes for competitors, to help in prefilling race forms.
{
"id": "http://example.org/profiles/sports/resources/tshirt",
"type": "AgentProfileResource",
"inScheme": "http://example.org/profiles/sports/v2",
"prefLabel": {
"en": "T-Shirt Preference"
},
"definition": {
"en": "Stores a t-shirt preference for prefilling race registrations."
},
"contentType": "application/json",
"inlineSchema": "{ \"type\": \"object\", \"properties\": { \
\"cut\": {\"enum\": [\"straight\", \"fitted\"], \"required\": true}, \
\"size\": {\"enum\": [\"x-small\", \"small\", \"medium\", \"large\", \
\"x-large\", \"2x-large\", \"3x-large\"], \"required\": true}}}"
}
This looks much like the placement
extension, with the addition of contentType
and without any recommended verbs.
These Concepts are just literal xAPI Activity definitions the Profile wants to provide for use. This is the Profile's canonical version of the Activity.
Property | Type | Description | Required |
---|---|---|---|
id |
IRI | The IRI of the activity | Required |
type |
String | Activity |
Required |
inScheme |
IRI | The IRI of the specific Profile version currently being described | Required |
deprecated |
Boolean | A boolean. If true, this Concept is deprecated. | Optional |
activityDefinition |
Object | An Activity Definition as in xAPI, plus a @context , as in the table below. |
Required |
Name | Values |
---|---|
@context |
SHOULD be https://w3id.org/xapi/profiles/activity-context and MUST contain this URI if array-valued. |
other properties | All as in xAPI 1.0.x Activity Definitions, defined at https://github.com/adlnet/xAPI-Spec/blob/master/xAPI-Data.md#activity-definition |
Except for @context
, the activityDefinition in this Concept MUST be a legal xAPI Activity Definition.
Profile Authors:
- MUST include a JSON-LD
@context
in all top-level objects of extensions, or in every top-level object if array-valued. This is due to restrictions in JSON-LD, and is not applicable to extensions with primitive values (string, number, boolean, null, or arrays thereof). - MUST ensure every extension
@context
is sufficient to guarantee all properties in the extension expand to absolute IRIs during JSON-LD processing.
Learning Record Providers using the Activity in Statements:
- MUST use the
id
for the Activityid
, and MUST NOT include@context
in the Activity Definition. - SHOULD either not include the definition or include all properties given here in the definition.
- if included, the properties SHOULD be exactly as given in the Profile, except for
name
anddescription
and other Language Maps. - Language Maps SHOULD only include languages appropriate to the situation.
- Language Maps MAY include languages not present in the Profile yet.
A Profile for competitive events might define Activities to represent standardized events, such as the 100 meter dash:
{
"id": "http://example.org/profiles/sports/activities/100mdash",
"type": "Activity",
"inScheme": "http://example.org/profiles/sports/v2",
"activityDefinition": {
"@context": "https://w3id.org/xapi/profiles/activity-context"
"type": "http://example.org/profiles/sports/activitytypes/event"
"name": {
"en": "100 Meter Dash"
},
"description": {
"en": "Represents the 100 meter dash as a general category of race. \
When this is in the 'grouping' contextActivities of a Statement, \
that means the Statement is about a particular 100 meter dash \
that is the object Activity."
}
}
}
The description
includes guidance on how to interpret this Activity's use in Statements, which also serves as guidance on how to use it.
A Statement Template describes one way Statements following the Profile may be structured.
Property | Type | Description | Required |
---|---|---|---|
id |
A URI for this Statement Template. | Required | |
type |
StatementTemplate |
Required | |
inScheme |
The IRI of the specific Profile version currently being described | Required | |
prefLabel |
Object | A Language Map of descriptive names for the Statement Template | Required |
definition |
Object | A Language Map of descriptions of the purpose and usage of the Statement Template | Required |
deprecated |
Boolean | A boolean, default false. If true, this Statement Template is deprecated. | Optional |
verb |
IRI | Verb IRI | Optional |
objectActivityType |
IRI | Object activity type IRI | Optional |
contextGroupingActivityType |
Array | Array of contextActivities grouping activity type IRIs | Optional |
contextParentActivityType |
Array | Array of contextActivities parent activity type IRIs | Optional |
contextOtherActivityType |
Array | Array of contextActivities other activity type IRIs | Optional |
contextCategoryActivityType |
Array | Array of contextActivities category activity type IRIs | Optional |
attachmentUsageType |
Array | Array of attachment usage type IRIs | Optional |
objectStatementRefTemplate |
Array | An array of Statement Template identifiers from this Profile version. | Optional |
contextStatementRefTemplate . |
Array | An array of Statement Template identifiers from this Profile version. | Optional |
rules |
Array | Array of Statement Template Rules | Optional |
A Statement Template MUST NOT have both objectStatementRefTemplate
and objectActivityType
.
The verb, object activity type, attachment usage types, and context activity types listed are called Determining Properties.
A Profile Author MUST change a Statement Template's id
between versions if any of the Determining Properties, StatementRef properties, or rules change. Changes of scopeNote
are not considered changes in rules.
A Learning Record Provider authoring a Statement following a Statement Template:
- MUST include all the Determining Properties in the Statement Template.
- MUST follow all rules in the Statement Template.
- MUST, if
objectStatementRefTemplate
is specified, set the Statement object to a StatementRef with theid
of a Statement matching at least one of the specified Statement Templates. - MUST, if
contextStatementRefTemplate
is specified, set the Statement context Statement property to a StatementRef with theid
of a Statement matching at least one of the specified Statement Templates.
A Profile Validator validating a Statement MUST validate all the Learning Record Provider requirements for a Statement Template are followed.
Statement Template Rules describe a location or locations within Statements using JSONPath, then describe the restrictions on the value(s) there, such as inclusion, exclusion, or specific values allowed or disallowed. For example, to require at least one grouping, the rules might be something like:
"rules": [
{
"location": "context.contextActivities.grouping[*].id",
"presence": "included"
}
]
They have these properties:
Property | Type | Description | Required |
---|---|---|---|
location |
String | A JSONPath string. This is evaluated on a Statement to find the evaluated values to apply the requirements in this rule to. All evaluated values from location are matchable values. |
Required |
selector |
String | A JSONPath string. If specified, this JSONPath is evaluated on each member of the evaluated values resulting from the location selector, and the resulting values become the evaluated values instead. If it returns nothing on a location, that represents an unmatchable value for that location, meaning all will fail, as will a presence of included . All other values returned are matchable values. |
Optional |
presence |
String | included , excluded , or recommended . |
Optional |
any |
Array | An array of values that needs to intersect with the matchable values. | Optional |
all |
Array | An array of values which all the evaluated values need to be from. | Optional |
none |
Array | An array of values that can't be in the matchable values. | Optional |
scopeNote |
Object | A Language Map describing usage details for the parts of Statements addressed by this rule. For example, a Profile with a rule requiring result.duration might provide guidance on how to calculate it. | Optional |
A Statement Template Rule MUST include one or more of presence
, any
, all
, or none
.
A Profile Author MUST include the keys of any non-primitive objects in any
, all
, and none
in additional @context
beyond the ones provided by this specification.
A Learning Record Provider authoring a Statement for the Statement Template including this Statement Template Rule:
- MUST include at least one matchable value if
presence
isincluded
- MUST NOT include any unmatchable values if
presence
isincluded
- MUST NOT include any matchable values if
presence
isexcluded
- MUST apply the following requirements if
presence
is missing, ifpresence
isincluded
, or ifpresence
isrecommended
and any matchable values are in the Statement:- MUST, if
any
is provided, include at least one value inany
as one of the matchable values - MUST, if
all
is provided, only include values inall
as matchable values - MUST NOT, if
all
is provided, include any unmatchable values - MUST NOT, if
none
is provided, include any values innone
as matchable values
- MUST, if
A Profile Validator validating Statements MUST validate the Statement Template Rule requirements for Learning Record Providers are followed. See the Communication document for further details on how to do so.
When validating a Statement for Statement Template Rules, contextActivities
normalization MUST have already been performed as described in the Experience API specification. That is, singleton objects MUST be replaced by arrays of length one.
The syntax and behavior of JSONPath is described at http://goessner.net/articles/JsonPath/index.html#e2. In addition, the following requirements, clarifications, and additions apply:
- Filter and script expressions MUST NOT be used.
- The union operator (a comma) may be used inside array or child expressions, so the result is the union on each expression being used separately.
- The legal values in an array or child expression are: strings (child expressions), non-negative integers (array expressions), the star character
*
representing all children/members, and unions of these as described above. - Any two or more legal JSONPath expressions, joined together by the pipe character
|
, optionally with whitespace around the pipe, are also considered a legal JSONPath expression. The value of this expression is all the values returned by each expression individually, flattened (that is, if one expression returns N values and another returns a single value, the combination returns N+1 values, not two values).
A Profile for competitive events might define a Statement Template for recording the outcomes of events:
{
"id": "http://example.org/profiles/sports/templates/placing",
"type": "StatementTemplate",
"inScheme": "http://example.org/profiles/sports/v2",
"prefLabel": {
"en": "Placing"
},
"definition": {
"en": "Records the actor placing in a particular event. \
The object is the specific event participated in, and an Activity \
representing the general type of event goes in grouping \
contextActivities."
},
"verb": "http://example.org/profiles/sports/verbs/placed",
"objectActivityType":
"http://example.org/profiles/sports/activitytypes/event",
"contextGroupingActivityType": [
"http://example.org/profiles/sports/activitytypes/event"],
"rules": [
{
"location":
"$.result.extensions['http://example.org/profiles/sports/extensions/place']",
"presence": "included"
}
]
}
Patterns describe groups of Statements matching particular Statement Templates, ordered in certain ways. For example, a Pattern in a video Profile might start with a Statement about playing a video and then be followed by Statements about pausing, skipping, playing again, and so forth.
Patterns have these properties:
Property | Type | Description | Required |
---|---|---|---|
id |
URI | A URI for the Pattern. | Required |
type |
String | Pattern |
Required |
primary |
Boolean | Default false. Only primary Patterns are checked for matching sequences of Statements. | Optional |
inScheme |
IRI | The IRI of the specific Profile version currently being described | Optional |
prefLabel |
Object | A Language Map of descriptive names for the Pattern | Optional |
definition |
Object | A Language Map of descriptions of the purpose and usage of the Pattern | Optional |
deprecated |
Boolean | A boolean. If true, this Pattern is deprecated. | Optional |
alternates |
Array | An array of Pattern or Statement Template identifiers. An alternates Pattern matches if any member of the array matches |
Optional |
optional |
URI | A single Pattern or Statement Template identifier. An optional Pattern matches if the identified thing matches once, or is not present at all |
Optional |
oneOrMore |
URI | A single Pattern or Statement Template identifier. A oneOrMore Pattern matches if the identified thing matches once, or any number of times more than once |
Optional |
sequence |
Array | An array of Pattern or Statement Template identifiers. A sequence Pattern matches if the identified things match in the order specified. |
Optional |
zeroOrMore |
URI | A single Pattern or Statement Template identifier. A zeroOrMore Pattern matches if the identified thing is not present or is present one or more times |
Optional |
A primary Pattern MUST include prefLabel
and definition
. They are optional otherwise.
A Pattern MUST contain exactly one of alternates
, optional
, oneOrMore
, sequence
, and zeroOrMore
.
A Profile Author MUST change a Pattern's id
between versions if any of alternates
, optional
, oneOrMore
, sequence
, or zeroOrMore
change. Note that if a Pattern used within another Pattern changes, the change will "bubble up" as each id
gets changed.
Profile Authors:
- MUST make sure their primary Patterns behave appropriately given the greedy matching rules in the algorithms section.
- MUST NOT put
optional
orzeroOrMore
directly inside alternates. - MUST NOT include any Pattern within itself, or within any Pattern within itself, or at any depth.
- MUST include at least two members in
alternates
. - MUST include at least two members in
sequence
, unlesssequence
is in a primary Pattern that is not used elsewhere and the member ofsequence
is a single Statement Template. - MAY re-use Statement Templates from other Profiles in Patterns. In this case, validating is done as if the Statement Template were present in this Profile.
- MAY re-use Patterns from other Profiles in Patterns. In this case, validating is done as if the Pattern were present in this Profile.
Learning Record Providers:
- MUST follow a Pattern from a Profile for every Statement that has the Profile in the category context activities.
- MUST send Statements following a Pattern ordered by Statement timestamp.
- MUST give Statements following a Pattern different timestamps if they are in different batches.
- SHOULD give all Statements following a Pattern different timestamps.
- MUST order Statements following a Pattern within the same batch with the same timestamp so ones intended to match earlier in the Pattern are earlier in the array than ones intended to match later in the Pattern
- MUST follow the rules given for a primary Pattern within a registration and possibly subregistration (a new extension). Specifically,
- all Statements following a primary Pattern MUST use the same registration.
- when only one Profile or only one Pattern occurrence of each Profile will be used, in a given registration, subregistrations SHOULD NOT be used.
- if any Statements following a primary Pattern do not contain a subregistration, all Statements with the same registration and Profile version in category MUST follow that primary Pattern.
- if any Statements with a given registration contain the subregistration extension, then all Statements with that registration with the same subregistration identifier for a given Profile version MUST follow the same primary Pattern.
- the extension key of the subregistration extension is https://w3id.org/xapi/profiles/extensions/subregistration.
- the subregistration extension MUST only be present in Statements with a registration.
- the subregistration extension is an array-valued context extension. The array MUST NOT be empty. Each value of the array MUST be an object with the properties in the table below.
Profile Validators validating Statements MUST validate all the requirements for Learning Record Providers for Patterns have been followed. See the Communication document for further details on how to do so.
Name | Values |
---|---|
profile |
The URI of a Profile present in the category context activities that this is a subregistration for. |
subregistration |
A variant 2 UUID as specified in RFC 4122. This is the subregistration identifier in the requirements above. |
Some Profiles may contain Patterns very similar to Statement data sent by previously existing Learning Record Providers, not strictly following this specification. It may be very close, but not follow it in all particulars, such as by missing a registration. While the details of how to handle this are outside the scope of this specification, Profiles aware of such existing data SHOULD make note of this and include descriptive language covering the degree of adherence.
A Profile for competitive events could define a primary Pattern and other associated Patterns for recording the running of a relay race.
{
"id": "http://example.org/profiles/sports/patterns/relay",
"type": "Pattern",
"primary": true,
"inScheme": "http://example.org/profiles/sports/v2",
"prefLabel": {
"en": "Relay Race"
},
"definition": {
"en": "A detailed recording of a relay race by a team in track and \
field. The actor of the `start` Statement should be the single runner \
who starts, and the actor of the `placed` Statement \
should be a Group of the entire team."
},
"sequence": [
"http://example.org/profiles/sports/templates/start",
"http://example.org/profiles/sports/patterns/handoffs",
"http://example.org/profiles/sports/templates/placing"
]
}
The primary Pattern for a relay race, it says a sequence of things happens: first, there's a Statement matching the start
template (not defined in the examples), then there's a Pattern of handoffs
, as defined next, and then there's a Statement matching the placing
template (defined in an earlier example).
{
"id": "http://example.org/profiles/sports/patterns/handoffs",
"type": "Pattern",
"inScheme": "http://example.org/profiles/sports/v2",
"definition": "The actor of each `handoff` Statement should be the runner \
receiving the baton, as described in that Statement Template.",
"oneOrMore": "http://example.org/profiles/sports/templates/handoff"
}
The Pattern for handoffs
is pretty simple--there's one or more of Statements matching the handoff
template. There's a definition on this Pattern even though it isn't required since it isn't primary, to aid people in using it.
In a real Profile these Patterns would be more complicated, accounting for various disqualifications, for example.
A common sort of occurrence will be the possibility of someone taking a "redo" of some section of an experience. For example, a Profile for a game supporting "redo" of levels after doing them the first time might have a Pattern like the following:
{
"id": "http://example.org/profiles/examplegame/patterns/levels",
"type": "Pattern",
"inScheme": "http://example.org/profiles/examplegame/v1",
"sequence": [
"http://example.org/profiles/examplegame/patterns/level1",
"http://example.org/profiles/examplegame/patterns/level1redos",
"http://example.org/profiles/examplegame/patterns/level2",
"http://example.org/profiles/examplegame/patterns/uptolevel2redos",
"http://example.org/profiles/examplegame/patterns/level3",
"http://example.org/profiles/examplegame/patterns/uptolevel3redos"
]
}
The levels
Pattern shows both the progression of levels and the possible redos between levels.
{
"id": "http://example.org/profiles/examplegame/patterns/level1redos",
"type": "Pattern",
"inScheme": "http://example.org/profiles/examplegame/v1",
"zeroOrMore": "http://example.org/profiles/examplegame/patterns/level1"
}
Level 1 redos are simple--the only level that can be redone is level 1.
{
"id": "http://example.org/profiles/examplegame/patterns/uptolevel2redos",
"type": "Pattern",
"inScheme": "http://example.org/profiles/examplegame/v1",
"zeroOrMore": "http://example.org/profiles/examplegame/patterns/uptolevel2"
}
For level 2, redos get more complicated, since someone can redo either level 1 or level 2.
{
"id": "http://example.org/profiles/examplegame/patterns/uptolevel2",
"type": "Pattern",
"inScheme": "http://example.org/profiles/examplegame/v1",
"alternates": [
"http://example.org/profiles/examplegame/patterns/level1",
"http://example.org/profiles/examplegame/patterns/level2"
]
}
The option of doing level 1 or level 2 with each redo is encapsulated in an alternates
Pattern.
The way JSON-LD documents are mapped onto semantics is through what's called a context, which is specified with @context
. Most of the time Profile authors and consumers do not need to worry about this at all -- this specification says what needs to go where, and provides the values to put in @context
in the necessary places. In addition to being hosted at the given URLs, the contexts used are also in the repository this specification is developed in.