Please note that the Service extension is still work in progress *this extension md file was previously named “SoftwareExtension”

SKG-IF Service Extension

A Service is a type of software application or component that provides specific functionality or operations over a network, often via the internet, and is typically accessed through an interface such as an API or a web application. (note Service is renaming of Software Service, which is still used in many discussion documents)

Properties

The properties set was suggested and informed from a number of examples both thematic and general:

NOTE property naming: properties introduced by this extension, that are not defined by authorative external origin such as Schema.org or introduced in the SKG-IF core entities are prefixed with “srv_” to avoid clashes with properties with the same name potententially introduced by other extensions. However the practice seems overkill in view of the checks automatically done by json parsing, especially in cases where the extension can be expected to be authorative. TBD.

Extensions to core entities

This extension adds new types to existing SKG-IF core entities:

Organisation

New types for modeling service operations:

  • hosting_organisation - organisation responsible for hosting and operating a service
  • research_infrastructure - organisation providing facilities, resources and services for research communities

Venue

New type for portal functionality:

  • srv_portal - a portal/catalogue through which services are discoverable and accessible

Properties

local_identifier

String (mandatory): Unique code identifying a [Software Service] in the SKG (if any, otherwise “stateless identifier”).

Suggestion: Use a URL as a string to make this entity dereferenceable on the Web. For additional information, see the section ‘Local identifiers of entities’ of the Interoperability Framework. A local_identifier can also be (derived) from an existing persistent identifier.

    "local_identifier": "11234/1-4816"

identifiers

List (recommended): Objects representing external identifiers for the entity.

Each identifier is structured as follows:

  • scheme String (mandatory): The scheme for the external identifier.
  • value String (mandatory): The external identifier.

Note: the current version of SKG-IF includes the following types of identifiers (to be specified as strings in the field “scheme”): doi, handle, pmid, url, omid, …

    "identifiers": [
        {
            "scheme": "handle",
            "value": "11234/1-4816"
        },
        {
            "scheme": "doi",
            "value": "10.18653/v1/K18-2020"
        }
    ]

entity_type

String (mandatory): Field stating what kind of entity is being serialised.

Needed for parsing purposes; fixed to service.

    "entity_type": "service"

name

Object (optional): The names of a [Software service] (multiple for multilingualism).

The object is a dictionary, the keys represent language codes following ISO 639-1; the special key none is reserved whenever the information about the language is not available or cannot be shared.

    "name": {
        "en": "UDPipe",
        "zh-cn": "UDPipe 工具"        
    },

other_names

Object (optional): The other names of a [Software service].

    "other_names": {
        "en": ["UDPipe 2", "UDPipe 2.0"]
    }

descriptions

Object (optional): The descriptions of a [Service] (multiple for multilingualism).

The object is a dictionary, the keys represent language codes following ISO 639-1; the special key none is reserved whenever the informtion about the language is not available or cannot be shared.

    "descriptions": {
        "en": ["UDPipe 2 is a Python prototype, capable of performing tagging, lemmatization and syntactic analysis of CoNLL-U input. It took part in several competitions, reaching excellent results in all of them", "Summary"],
        "cs": ["UDPipe je trénovatelný nástroj pro tokenizaci, tagging, lemmatizaci a závislostní parsing CoNLL-U souborů. UDPipe je jazykově nezávislý a pro natrénování modelů využívá anotovaná data ve formátu Universal Dependencies"],
        "none": ["ontaligestring"]
    }

srv_audience_byrole

Object (optional): The audience(s) that the service is intended to be used by This can both express desire and/or design of the service operators. Values are mandatory taken from the https://vocabs.sshopencloud.eu/vocabularies/sshoc-audience/audienceScheme

    "@context": {
        "sshocaudience": "https://vocabs.sshopencloud.eu/vocabularies/sshoc-audience/"
    },
    "srv_audience_byrole": ["sshocaudience:public", "sshocaudience:student" ]

srv_audience_byjurisdiction

Object (optional): The jurisdiction that is given by the service operator’s legal status limits. the audience. Values taken from either Global, Institution, National, or Regional aka multiple countries, from (https://zenodo.org/records/15516020).

    "srv_audience_byjurisdiction": ["Institution", "National" ]

disciplines

List (optional): The disciplines to which a [Software Service] is dedicated. The disciplines must be specified using the Library of Congress Classification codes, available at https://id.loc.gov/authorities/classification (e.g. PA3000-PA3049 for classical literature). In case a [Service] is discipline agnostic, the string “all” should be specified.

    "@context": {
        "loc": "https://id.loc.gov/authorities/classification/"
    },
    "disciplines": [
        "loc:QC790.95-QC791.8"
    ]

    "@context": {
        "loc": "https://id.loc.gov/authorities/classification/"
    },
    "disciplines": ["all"]

is_accessible_for_free

Boolean (optional): A property to signal that the Service is accessible for free.


    "is_accessible_for_free": true

srv_invocation_type

List (mandatory): The way the service is used or called. Multiple values are possible, access rights and licenses are assumed to be the same. Values are specified by vocabulary: https://vocabs.sshopencloud.eu/vocabularies/invocation-type/invocationTypeScheme

    "@context": {
        "sshocinvt": "https://vocabs.sshopencloud.eu/vocabularies/invocation-type/"
    },
    "srv_invocation_type": [ "sshocinvt:restfullWebservice", "sshocinvt:webApplication" ]

srv_life_cycle_status

List (optional): Indicates the development cycle and/or maturity status of the service. Values are by vocabulary: https://vocabs.sshopencloud.eu/vocabularies/eosc-life-cycle-status/ Originally specified in the EOSC Service Profile. Could be extended with TRL classifications.

     "@context": {
        "elcs": "https://vocabs.sshopencloud.eu/vocabularies/eosc-life-cycle-status/"
     },
    "srv_life_cycle_status": ["elcs:life_cycle_status_production", "elcs:TRL6" ]

srv_availability_geographic

List (optional): list of countries and regions where the service is made available, eg. for license reasons. Values are by the vocabulary: https://vocabs.sshopencloud.eu/vocabularies/eosc-geographical-availability/

     "@context": {
        "eoscgeoavail": "https://vocabs.sshopencloud.eu/vocabularies/eosc-geographical-availability/"

     },
    "srv_availability_geographic": ["eoscgeoavail:eu","eoscgeoavail:uk"]

website

String (mandatory): Landingpage for the service. Preferably one maintained by the service operator

    "website": "https://ufal.mff.cuni.cz/udpipe/2"

srv_supported_language

List (optional) if applicable the language(s) the service is able to process, values provided as ISO369-3 language codes using Lexvo.org published vocabulary NOTE properties introduced by this extension, that are not from external origin or also used in core entities are prefixed with “srv_” to avoid clashes with properties with the same name potententially introduced by other extensions

    "@context": {
       "lexvo-iso639-3": "http://lexvo.org/id/iso639-3"
     },
     "srv_supported_language": ["lexvo-iso639-3/deu","lexvo-iso639-3/nld"]

topics

List (optional): [Topic] object identifiers relevant for the scope (topic) of a [Service]. NOTE this description is not consistent with other examples, should be discussed and repaired.

    "topics":  [
    {
            "term": "topic_1",
            "provenance": [
                {
                    "associated_with": "openaire-infra",
                    "trust": 0.7
                }
            ]
    } ]

srv_research_infrastructure

List (optional): Is associated with an [Organisation] that provides facilities, resources and services for the research communities to conduct research. NOTE When querying the SKG-IF API, query responses may return as value for this property the complete Organisation entity information, but should minimally return local_identifier and entity_type.

     "srv_research_infrastructure": [
       {
         "local_identifier": "https://ror.org/03wp25384",
         "entity_type": "organisation",
         "name": "CLARIN ERIC",
         "types": [
           "facility",
           "research_infrastructure"
         ],
         "country": "NL"
       },
       {
         "local_identifier": "https://ror.org/03wp25384",
         "entity_type": "organisation",
       }
     ]

srv_hosting_organisation

List (optional): Is depending on [Organisation] reponsible for hosting a service or infrastructure component. NOTE When querying the SKG-IF API, query responses may return as value for this property the complete Organisation entity information, but should minimally return local_identifier and entity_type.

     "srv_hosting_organisation": [
       {
         "local_identifier": "https://ror.org/00dd4fz34",
         "entity_type": "organisation",
         "name": "Digital Research Infrastructure for Language Technologies, Arts and Humanities",
         "short_name": "LINDAT",
         "types": [
           "facility",
           "hosting_organisation"
         ],
         "country": "CZ"
       },
       {
         "local_identifier": "https://ror.org/00dd4fz34", 
         "entity_type": "organisation"
       }
     ]

List (optional): Is the specific [Organisation] legally reponsible for the service operation and publishing NOTE When querying the SKG-IF API, query responses may return as value for this property the complete Organisation entity information, but should minimally return local_identifier and entity_type.

     "srv_hosting_legal_entity": [{
        "local_identifier": "https://ror.org/024d6js02",
        "entity_type": "organisation",
        "name": "Charles University",
        "types": [
          "education",
          "research"
        ],
        "country": "CZ"
     }]

relevant_organisations

List (optional): [Organisation] identifiers associated with and relevant for a [Service]. Identifiers can be of local or global identifier system type eg. ror, uri. Organizations that are identified by a local_identifier, can be given additional types: “research_infrastructure”, “hosting_organisation” and “hosting_legal_entity”. NOTE this is provided by adding new types to the Organisation entity types

	"relevant_organisations": ["https://ror.org/024d6js02", "https://ror.org/03wp25384"]

Object (optional): A dictionary of objects representing related [Research products], where the semantics of such relationships is specified as a key and the related product by its local_identifier. NOTE (TODO): THIS LIST HAS TO BE MODFIED FOR SERVICE/SOFTWARE RELATIONS It is structured as follows:

  • cites List (optional): [Research products] identifiers that are cited by a given Research product.
  • is_supplemented_by List (optional): [Research products] identifiers that are supplement of a given Research product.
  • is_documented_by List (optional): [Research products] identifiers that documents a given Research product.
  • is_new_version_of List (optional): [Research products] identifiers that are prior versions of a given Research product.
  • is_part_of List (optional): [Research products] identifiers that contain the current Research product.
    "related_products": {
        "cites": ["product_2", "product_3", "product_4"],
        "is_supplemented_by": ["product_7", "product_8", "product_9"],
        "is_documented_by": ["product_10", "product_13"],
        "is_new_version_of": ["product_10", "product_13"],
        "is_part_of": ["product_11"]
    }

keywords

List (optional): List of keywords relevant for service discovery, values may be simple strings or concept URIs.

    "keywords": ["https://www.wikidata.org/wiki/Q30642","parsing"]

srv_deployment_of

List (optional) The software that this service is a running instance of.

Can be:

  • A URI string (e.g., a source code repository link)
  • A typed reference:
    • skg:research_product — a Research Product of type software
    • fabio:Software — a software entity
    • schema:SoftwareSourceCode — a source code entity
  "srv_deployment_of": [                                                                            
      "https://github.com/ufal/udpipe",                                                             
      { "@id": "http://example.org/software/udpipe", "@type": "fabio:Software" },                   
      { "@id": "http://example.org/source/udpipe", "@type": "schema:SoftwareSourceCode" },          
      { "@id": "http://example.org/research_product/RP_101", "@type": "skg:research_product" }      
  ]

srv_contributions

List (optional) [Agents] that contributed to a [Service] (optional): A dictionary of objects representing contributing [Agents], where the semantics of the contributionis specified as a key. #TODO: THIS LIST OF CONTRIBUTION TYPES HAS TO BE MODFIED FOR SERVICE/SOFTWARE CONTRIBUTIONS #Recommended to use the specific properties such as “research_infrastructure”, “hosting_organisation” etc. to measure contribution of organisations. Individual person contributions of software developers, data managers etc. are well taken care under research-products. It is difficult to track individiual contributions to operations anyway

It is structured as follows:

    "srv_contributions": [ 
               {
                    "by": "University of Sheffield",
                    "role": "operator"
               },
               {
                    "by": "UK Research and Innovation agency",
                    "role": "funder"
               }  
    ]