euBusinessGraph ontology

NUTS2 (administrative region) of the address.

Use a valid NUTS region. Eg Trentino-Alto Adige/Sudtirol (taken from the Geonames hierarchy above) is not a NUTS region.

nuts:ITH2 Provincia Autonoma di Trento

nuts:ITC2 Valle d'Aosta/Vallée d'Aoste

NUTS3 (Province) of the address.

nuts:ITH31 Verona

nuts:ITC20 Aosta

LAU1 (Municipality/Commune/Settlement) of the address.

Some countries (e.g. Bulgaria) use both LAU1 and LAU2 levels. Others (e.g. Italy) use only LAU2.

LAU2 (Municipality/Commune/Settlement) of the address.

Some countries (e.g. Bulgaria) use both LAU1 and LAU2 levels. Others (e.g. Italy) use only LAU2.

Resolution of a geographic coordinate location: from continent (least precise) to individual house number (most precise).

Values must be mapped to the ConceptScheme (lookup list).

Web resource associated with an identifier system

Specifies a template that can be used uniformly to build URLs for all identifiers in the system.

'https://opencorporates.com/companies/{}.xml'  with MIME type 'application/xml'

'https://opencorporates.com/companies/{}'      with MIME type 'text/html'

'https://opencorporates.com/companies/{}.rdf'  with MIME type 'application/rdf+xml'

'https://opencorporates.com/companies/{}.json' with MIME type 'application/json'

URL providing human or machine-readable rule(s) for validating identifiers in the system

Can be in the form of webpage, PDF document, RDF shape, etc. Multiple values about the same validation rule are ok.

Web resource associated with an identifier system.

Agent that issues identifiers and then keeps them in a database (register), and who issued a specific identifier

In many countries there is a single registry although in others, such as Spain and Germany, multiple registries exist. [rov:].

Companies House is the issuer for the official UK identifier system.

Web URLs are not issued by any central agency and there is no register to consult.

Dun and Bradstreet issues DUNS numbers.

GLEI Foundation (GLEIF) issues identifiers for the world-wide GLEI register (through its Local Operating Units, LOU).

The Identifier System this identifier is part of

RSS/Atom feed pertaining to the company.

Agent who specified the rules and organisation of an Identifier System

Dun and Bradstreet specified DUNS numbers

W3C or TimBL specified web URLs

W3C specified web URL rules

GLEIF specified GLEI

GLEIF specified LEI structure and rules, then formalied it as ISO 17442:2012 Financial services -- Legal Entity Identifier (LEI)

Email that is officially registered and with the same validity as certified (snail) mail.

The value should be a mailto: URL.

mailto:elettorale@pec.comune.trento.it

Geographic coordinates, eg of a geo-located address.

If data is integrated from several providers, we may end up with multiple coordinate pairs in that node, which will need to be resolved into one.

License that applies to an identifier system or dataset

If possible, use established license URLs, e.g. from https://creativecommons.org/ or http://rightsstatements.org/

Wikipedia page pertaining to the company.

Website pertaining to the company or URL of a web resource

For system BG TR, <https://brra.bg> with language "bg" and MIME type "text/html" is the homepage

For system BG TR, <http://opendata.government.bg/dataset/tbprobckn-pernctbp> with language "bg" and MIME type "application/xml" is open data download in XML

For system BE Banque-Carrefour des Entreprises, <https://kbopub.economie.fgov.be/kbopub/zoeknummerform.html> with MIME type "application/x-search" is the search URL

Beginning of the membership relation

End of the membership relation

An identifier of a Company (the official identifier is rov:registration) or Identifier System / Register

This includes, but is not limited to: - Direct: Identifier in some aggregated dataset, where the official ID is part of the aggregation ID. - Indirect: Identifier in some aggregated dataset or for a website, where the aggregation ID bears no resemblance to the official ID - Multiple: Identifier that has a many-to-many relation with the company, i.e., a company may have several identifiers of the same kind, and the same identifier can be used by several companies.

Multiple: bank license, insurance license, license to trade dual-use goods (arms), Twitter account, Facebook account.

Indirect: GLEI LEI, SDATI Atoka ID.

Direct: EU VAT number (in those jurisdictions where it is used as the primary company identifier), BRIS id, OCORP id.

Indirect: BG GUID 617f4edf8c154f4296efdf146513de21, which corresponds to official id 204060254 and can be used to reach the official register page for that company

Country of the address

nuts:IT (Italy).

NUTS1 (Macroregion) of the address.

nuts:ITC NORD-OVEST (Nord-est).

Public legal address where legal papers can be served.

org:hasRegisteredSite leads to a node with types org:Site and locn:Address that has self-link org:siteAddress.

Other address/location associated to an organisation.

org:hasSite leads to a node with types org:Site & locn:Address that has self-link org:siteAddress. org:hasRegisteredSite is declared a sub-property of org:hasSite, so querying by org:hasSite will obtain all addresses.

The officer that the membership is about

Interval (beginning and end) of the membership relation. May be an open interval.

for the common case of missing time, we prefer to use xsd:date rather than completing it with a fake null time, eg "T00:00:00Z"^^xsd:dateTime.

Points to the company the membership is about

The role played by the officer in the membership

EBG doesn't need the distinction between org:Site and locn:Address, so we use one node having both types, and a self-link org:siteAddress

Economic activity of the organization (NACE code)

Recorded using a controlled vocabulary: EC NACE 2. Values should be like nace:A, nace:42.50, etc

Flag that identifies whether a company is active or not [ebg:]. The operational and/or legal registration status of the entity [lei:].

A Best Practice for recording status levels is to use the relevant jurisdiction's terms and to encode these in a SKOS Concept Scheme [rov:].

'insolvent', 'bankrupt', 'in receivership': likely to mean slightly different things with different legal implications in different jurisdictions [rov:].

'Actively trading' vs 'Dormant' vs 'Unregistered' could be considered the 3 main divisions of Status.

'Normal Activity': does appear to have cross-border usefulness and this should be used in preference to terms like 'trading' or 'operating' [rov:].

Company Type (Legal Form of the entity)

In addition or alternatively to this field, ebg:orgTypeText may be used.

Identifier that holds the official company registration in its jurisdiction of registration. It establishes the legal existence of the company.

This is a sub-property of adms:identifier, and should be an adms:Identifier pointing to an official ebg:IdentifierSystem. So it holds redundant information, but this is done for compatibility with the Registered Organization vocabulary.

Year the organization was dissolved or deregistered, as integer.

Used for faceting.

Exclusion note of a NACE classification concept

Year the organization was created, as integer.

Used for faceting.

Whether the system has an Issuer and issued identifiers are kept in a database (register)

Websites are not enumerated

Every official register is enumerated

whether identifiers can change

Some registers encode additional information within the identifier (e.g., some identifiers encode the economic sector of the company). In these cases the identifier changes when the extra information that is encoded changes (see 'Dumb vs. Intelligent Identifiers' in Section 3.1.2.5).

Whether the system is considered the official one in all jurisdictions in which it applies.

True for UK Company House identifiers, the official register for the GB jurisdiction

False for Website, Twitter, Facebook in that they don't have any official status

False for SDATI identifiers since the SDATI system is not official in Italy.

False for GLEI, since it is not the official register for any of the world's jurisdictions

Whether identifiers can be removed from the register (e.g., when a company is dissolved): such are not persistent

Some registers 'remove' identifiers when the object they relate to is no longer active, whereas the best practice is to make such identifiers Obsolete.

whether identifiers from the system are available for public use: consulting, search or download.

Most public systems (but not all) will have some Publishers, and preferably some web resources.

DUNS identifiers are not public, even though there are limited lookup services (e.g., https://www.dnb.com/duns-number/lookup.html). Dun and Bradstreet does not allow their use en-masse.

Italian (IT) company identifiers are public. Even though they are not published openly, they are available for a fee, and then can be used freely

Whether the company is publicly traded (listed at a stock exchange) [ebg:].

Whether each entity has only one identifier in the system

Similar to owl:FunctionalProperty, and Wikidata Single Value constraint

Official person names are single-valued, since each person has one official name (but that can change over time).

Stock exchange tickers are not single-valued, since a company may have several tickers for different kinds of its stock.

Websites are not single-valued since a company may have several websites

Company IDs are single-valued in most national registers. If two companies merge, the recessive company ID is usually deprecated.

Whether the company is a startup [ebg:]

A special section of the IT Business Register http://startup.registroimprese.it/ includes pitch, sectors and business model of startups and innovative SMEs

Whether this organisation is owned by the government, a government agency, community, city or other public entity.

In many cases it's not possible to compute this attribute without access to the shareholder register, so it may be missing.

True for the Autonomous Province of Trento (local government).

False for SpazioDati (a private company).

True for the Norwegian company Statkraft (Norwegian state-owned company).

Whether each identifier in the system relates to only one entity

Similar to: owl:InverseFunctionalProperty, and Wikidata Distinct Values constraint

Person names are not unique since they are highly ambiguous

Company IDs are unique in most national registers: one ID cannot identify two companies

Websites are not unique since a website can be shared by several related companies

Stock exchange tickers are unique to each exchange (but a company can buy the ticker of another, i.e. the identity may change over time)

Hierarchical level of a thesaurus concept

The number of employees in the company.

Global order of a thesaurus concept respecting the hierarchical structure (depth-first traversal)

Economic activity of the organization (free text)

Such field is widely used in some countries (e.g. BG) and may include much richer info than rov:orgActivity

NACE = 72.19 Scientific research and development in the field of natural, medical, agricultural and technical sciences, without biotechnology

Free text = Development, exploration and production of opto-electronic systems for analogue and digital holographic recording, interferometric measurement of mechanical characteristics and biological objects at micro and macro level, refractometers, light-sensitive materials for holographic recording, blueprints of museum exhibits, artworks and collection surveys, production of holographic illustrations for students, albums, advertising, souvenirs and others, design and manufacture of holographic optical elements with a wide spectrum of applications, marketing and sale of holographic products, consulting, training and promotion of holographic methods and technologies, mediation, brokerage, and **any other activity not prohibited** by Bulgarian legislation.

Company status as it comes from the original register (free text).

If both Status and Status Text are present, they must express compatible values (Status Text may be a finer-granularity value).

Inactive

Dissolved

Revoked

Company Type (Entity Legal Form) as reported by a national company registe (free text).

Include valid language tag (see [iana:] for list).

Private Limited Company

GLEI RAL code for the identifier system.

Used only for official registers.

Pattern to use together with the Validation Regex to normalise identifier values by removing optional decorations

"$1$2$3" can be used together with the validationRegex "(\d{2})-?(\d{3})-?(\d{4})" to extract the pure digits from a DUNS number spelled with optional dashes (e.g. "36-032-1459" -> "360321459")

The (hierarchical) position of the person within a company (in free text), e.g., director.

URL template of an identifier web resource. If it has a placeholder {}, substitute the identifier value there. If it has placeholders like $1, $2… substitute the groups extracted by ebg:validationRegex

'For BR, 'http://data.brreg.no/enhetsregisteret/enhet/{}' with either 'application/xml' or 'application/json' returns the respective MIME type using Content Negotiation'

For OCORP, 'https://opencorporates.com/companies/{}.json' returns company data with MIME type 'application/json'

For OCORP, 'https://opencorporates.com/companies/{}'      returns company data with MIME type 'text/html'

For OCORP, 'https://opencorporates.com/companies/{}.xml'  returns company data with MIME type 'application/xml'

'For the EU Value Added Tax (VIES) register, 'http://ec.europa.eu/taxation_customs/vies/vatResponse.html?memberStateCode=$1&number=$2' with MIME type 'text/html' shows a web page for an identifier that can be parsed with ebg:validationRegex '([A-Z]{2})/(\d+)'. Here $1 is the member state code, and $2 is the national company id'

For OCORP, 'https://opencorporates.com/companies/{}.rdf'  returns company data with MIME type 'application/rdf+xml'

'For BE Banque-Carrefour des Entreprises, 'https://kbopub.economie.fgov.be/kbopub/toonondernemingps.html?ondernemingsnummer={}' with MIME type 'text/html' shows a web page for the company.'

Regular expression for validating identifier values within the system

"(\d{2})-?(\d{3})-?(\d{4})" is a regex for validating DUNS numbers that may include optional dashes in the indicated positions (e.g. "36-032-1459")

"\d{9}" is a simple regex for validating DUNS numbers

"([A-Z]{2})/(\d+)" is a regex for the EU Value Added Tax (VIES) register, which consists of a member state code followed by slash and a national numeric identifier

Jurisdiction in which the company is registered or to which the identifier system applies.

For Europe, this is a country, even for Germany that has 166 separate registers (see Registration Authorities List). Many jurisdictions register foreign companies that are incorporated and fully owned in another country. EBG collects all registrations from each register and doesn't attempt to match foreign companies across jurisdictions.

BG has many types, including a number of European companies.

NO has type NUF 'Norskregistrert utenlandsk foretak'.

MIME type of the resource. If several are provided, the server must provide all these resource types using Content Negotiation

'For a search page, use the custom type 'application/x-search''

text/csv

text/tab-separated-values

'For a Description document supporting the OpenSearch standard, use '<http://a9.com/-/spec/opensearch/1.1/>''

application/rdf+xml

application/xml

text/html

text/turtle

When was the identifier was issued (start date).

This may or may not be the same as the Incorporation Date of the company. Even for the official registration, the two dates may differ, depending on business rules.

NUTS/LAU Code of an administrative region

Level of an administrative region: NUTS1..3=1..3, LAU1..2=4..5

Name of an administrative region or settlement

Languages used by the organization.

Currently this data is provided by SDATI, for matching company entities only.

en

zh-CN

it

The birth date of the person.

Description of the resource

Date when the organization was dissolved or removed from register.

Should have a valid XSD datatype.

2010-11-18

2010-11-01

2010-01-01

Date when the identifier expires (end date)

This may or may not be the same as the dissolution date of the company. Even when it is about the official registration, the two dates may differ, depending on business rules.

Date when the organization was created.

Should have a valid XSD datatype.

2010-11-18

2010-11-01

2010-01-01

The gender of the person.

An identifier for a resource.

Used to represent the identifier (e.g. URL) of a Person or Organization.

Language of the web resource

Use a valid [iana:] language code from the IETF BCP 47 standard

Coordinates must be expressed in the WGS 84 system.

If no coordinates are available, do not make a schema:GeoCoordinates node.

Name or short (generic) description of the resource

Provide this since it's useful to the user. Can provide it in several languages, in which case the language tag should be specified

Homepage

Company data

Search

Download

The nationality of the person.

Agent that publishes an identifier database (register) in some form (i.e., printed form, online)

Different functions and distributions can be considered (e.g. search or other query, per-company web resources, full download). Often but not always the issuer is also the publisher.

GLEIF publishes the GLEI openly, with a full dump that is updated regularly, which has spurned re-publishers such as http://openleis.com/ and http://glei.info.

DUNS numbers are not openly published.

Companies House and BRC publish the UK and NO registers openly.

Informal/popular name of the company (also called Trading As) [ebg:]. Sometimes (e.g., when the legal name is very long) people refer to a company using a different, informal version [ebg:]. Some jurisdictions recognize concepts such as a trading name or alternative forms of a legal entity's name [rov:].

Do not emit such name if it's equal to the legal name [ebg:].

OpenCorporates

Literal value of the identifier.

adms:Identifier recommends that this value should be "datatyped with the identifier scheme (including the version number if appropriate)". However, we decided not to do this because: - One cannot search by exact literal value unless one also knows the datatype URL - Currently we do not plan to maintain different identifier system versions

A single preferred name of a company or identifier system (preferably English).

Used as a display name of the company or identifier system. For companies we usually pick one of the Registered Names (English if present, else at random). For identifier systems, we use a commonly used English name.

Position of an instant, expressed using xsd:date.

Part of a city, village or neighborhood.

Not likely to appear, but we have Coordinate resolution corresponding to this field.

Full address, free text.

Aston House, Cornwall Avenue, London N3 1LF, UK

Via Belenzani 19, 38122, Trento

Street number and/or building name.

Not likely to appear separately, but we have Coordinate resolution corresponding to this field.

47A

Polygraphia Office Center, floor 4

Some addresses are associated with a PO box instead of a street address.

Postal code of the address.

If possible, a European postal code should correspond to the fixed values given in NUTS Downloads, so we can correlate it to a NUTS region.

Locality/City/Settlement of the address, free text.

Provide a language tag if possible. If both ebg:adminUnitL6 and locn:postName are specified, they should correspond to each other.

Street name and optionally number.

It is ok if the street number is also here, because many systems do not enter the number separately. There may be several concatenated lines, separated with newlines.

Via Belenzani 19

Polygraphia Office Center, floor 4
47A Tsarigradsko Shosse

The legal name of the person which is usually determined at or around birth, usually persistent and for this reason usually recorded by some official registers. Although birth names tend to be persistent we do not assume this has to be the case.

Michele Barbera

The legal name of the business, i.e. official name of the company [ebg:]

A business may have more than one legal name, particularly in jurisdictions with more than one official language (e.g. Belgium). Some registries also treat a transliterated name as official [ebg:]

Ontotext

CHRINON LTD