Japan Link Center (JaLC)

This section describes how the OpenAIRE Graph integrates the metadata of the research products registered with the Japan Link Center (JaLC), the DOI Registration Agency of Japan. It covers where the data comes from, how it is acquired and transformed, and how the JaLC metadata fields populate the fields of the OpenAIRE Graph data model.

Data source

JaLC is the DOI Registration Agency of Japan, jointly operated by the Japan Science and Technology Agency (JST), the National Institute for Materials Science (NIMS), the National Institute of Informatics (NII) and the National Diet Library (NDL). It registers DOIs for Japanese scholarly content. JaLC members can also register their DOIs with Crossref or DataCite through JaLC.

Each JaLC record declares its content type in the field content_type:

content_type	Content
JA	Journal articles
BK	Books and reports
RD	Research data
EL	E-learning materials
GD	General data and documents

Data acquisition

As part of its data provision services (in Japanese), JaLC distributes the metadata of all its DOIs as bulk dump files in JSON format. The dump is updated every six months. According to the JaLC Terms of Service (in Japanese), DOIs, bibliographic data and URLs can be used freely, regardless of purpose. The dump does not include abstracts.

The dump comes in three packages, one per registration agency of the DOIs. The August 2025 dump contained:

Package	DOIs
DOIs registered with JaLC	9,847,719
DOIs registered with Crossref via JaLC	2,585,672
DOIs registered with DataCite via JaLC	6,079

Each package holds one archive per DOI prefix, which in turn holds one JSON document per DOI.

Every collection imports a complete dump, so the JaLC records in the Graph reflect the dump available at collection time.

Processing workflow

1. Download: the three packages of the dump are downloaded from JaLC.
2. Conversion: every JSON document becomes one XML record. The conversion only changes the syntax: each JSON field becomes an XML element with the same value, and each list becomes a repeated element. The record header carries the DOI (as doi:<DOI>), the last update date of the JSON document (updated_date) and the registration agency (ra). Documents without a DOI are skipped, and the few characters that XML 1.0 does not allow are removed.
3. Collection: the OpenAIRE aggregation system collects the resulting file for the data source Japan Link Center, one record per DOI.
4. Transformation: the transformation rule xslt_cleaning_JaLC_dump_to_hadoop maps each record into the OpenAIRE internal metadata format, which is based on the DataCite schema. The mapping choices described below are made at this step.
5. Graph production: each transformed record becomes a research product with its relations, which then go through the standard steps of the Graph production workflow (cleaning, deduplication, enrichment).

Record filtering

A record is not transformed when:

• its content_type is not one of JA, BK, RD, EL, GD;
• it has no doi.

Entity Mapping

The table below describes how the fields of a JaLC record populate the fields of an OpenAIRE research product. JSON paths are relative to the data object of the JSON document; [] marks a list.

OpenAIRE Research Product field path	JaLC record JSON path	Notes
Research product
id	doi	id in the form efb2be5d8919::md5(doi:<doi>), built from the local identifier, because JaLC is not a PID authority for DOIs in the Graph
originalIds	doi	doi:<doi> and the internal identifier
type	content_type, resource_type	Using the dnet:result_typologies vocabulary, the instances.type is resolved to one of publication, dataset, software, other; see Research product types
mainTitle	title_list[].title	JaLC provides one title per language (title_list[].lang): every language version is kept as a main title, and one of them is shown as mainTitle
subTitle	title_list[].subtitle
authors	creator_list[]	One author per creator, in the order of the source
publicationDate	publication_date.publication_year, publication_date.publication_month, publication_date.publication_day, or date	See Publication dates
publisher	publisher_list[].publisher_name	The English name (publisher_list[].lang = en) when available, otherwise the first one
language	content_language, or journal_txt_lang when content_language is missing	cleaned by using vocabulary dnet:languages
subjects	keyword_list[].keyword	scheme=keyword; one subject per keyword and language
bestAccessRight		Derived from instances.accessRight
collectedfrom	(fixed value)	The data source Japan Link Center, openaire____::4965150f44bea971d2749c39dfdcacaf, for the records of all three packages. The same data source is set as the host of the instance
organizations	creator_list[].affiliation_list[].affiliation_identifier_list[]	Only identifiers whose type is ROR; see Relation Mapping
Author Mapping
authors.fullName	creator_list[].names[].last_name, creator_list[].names[].first_name	For persons, last_name, first_name, taken from the English name (names[].lang = en) when available, otherwise from the first name in the list. For organisations (creator_list[].type is institute or corporate), the organisation name
authors.name	creator_list[].names[].first_name	Persons only; for organisations the Graph may derive it from fullName
authors.surname	creator_list[].names[].last_name	Persons only; for organisations the Graph may derive it from fullName
authors.rank		Sequential number starting from 1
authors.pid	creator_list[].researcher_id_list[].id_code	Only when researcher_id_list[].type is ORCID, persons only; scheme = orcid_pending
Container Mapping		Journal articles (content_type = JA) only
container.name	journal_title_name_list[].journal_title_name	The full English title when available, then any full title, then the English one, then the first one
container.issnPrinted	journal_id_list[].journal_id	Where type = ISSN and issn_type = print
container.issnOnline	journal_id_list[].journal_id	Where type = ISSN and issn_type = online. The linking ISSN (issn_type = issn-l) does not populate container.issnLinking
container.vol	volume
container.iss	issue
container.sp	first_page
container.ep	last_page
Instance Mapping		One instance per record
instances.type	content_type, resource_type	A term of the dnet:publication_resource vocabulary; see Research product types
instances.alternateIdentifiers	doi, alternate_identifier_list[]	The DOI, with scheme = doi. JaLC is not a PID authority for DOIs in the Graph, so the DOI is an alternate identifier, not a value of instances.pids. From alternate_identifier_list[], only the types PMID, BIBCODE and URN are mapped
instances.urls	doi, url	The DOI URL https://doi.org/<doi>. url is added only when it differs from the DOI URL; in the dump it is normally the DOI URL itself
instances.accessRight	content_type, relation_list[]	See Access rights
instances.publicationDate	as publicationDate
instances.refereed	content_type	peerReviewed for journal articles (JA). The source gives no review information for the other content types, which are set to nonPeerReviewed by the Graph cleaning

All the records are related to the data source with name=Japan Link Center and id=openaire____::4965150f44bea971d2749c39dfdcacaf.

Research product types

The instance type depends on content_type and, for research data, on resource_type (the DataCite resourceTypeGeneral). The research product type then follows from the instance type through the dnet:result_typologies vocabulary.

content_type	Condition	instances.type	type
JA		Article	publication
BK	an alternate_identifier_list[] value of type JAEA-REPORT or JAERI-REPORT containing conf	Conference object	publication
BK	any other value of type JAEA-REPORT or JAERI-REPORT	Report	publication
BK	otherwise	Book	publication
GD	the publisher name contains Jxiv, or site_name contains preprint	Preprint	publication
GD	otherwise	Other literature type	publication
EL		Lecture	other
RD	resource_type = Image	Image	other
RD	resource_type = Audiovisual	Audiovisual	other
RD	resource_type = Software	Software	software
RD	resource_type = Text	Other literature type	publication
RD	resource_type = Collection	Collection	dataset
RD	resource_type = Event or Other	Other ORP type	other
RD	resource_type = PhysicalObject	PhysicalObject	other
RD	resource_type = Dataset, or no resource_type	Dataset	dataset

Access rights

The access right of the instance depends on the content type and on the presence of a link to the full text:

• OPEN (COAR c_abf2) for books (BK);
• OPEN for the other content types when relation_list[] contains a link to the full text PDF (relation_list[].relation = fullTextPdf);
• UNKNOWN otherwise.

Publication dates

publicationDate is built from publication_date: the full date when year, month and day form a valid date, otherwise the first day of the month, otherwise the first day of the year. When publication_year is missing, the flat field date is used instead.

JaLC uses placeholder years (such as 9999, 0000 and 1000) for records whose publication date is unknown; they concern about 5% of the records. To avoid publishing such values as dates, a publication year is accepted only between 1300 and the next calendar year. Outside this window the research product has no publication date.

Relation Mapping

OpenAIRE Relation Semantic and inverse	JaLC record JSON path	Source/Target type	Notes
Cites/IsCitedBy	citation_list[].doi	ResearchProduct/ResearchProduct	The target is identified by the cited DOI, as doi_________::md5(doi). References given only as free text (citation_list[].original_text) are not mapped
hasAuthorInstitution/isAuthorInstitutionOf	creator_list[].affiliation_list[].affiliation_identifier_list[]	ResearchProduct/Organization	Only identifiers whose type is ROR; the target is the ROR organization

See Relationship types for the semantics of the relations.

Fields not mapped

The following JaLC fields do not populate the Graph: funding information (fund_list), descriptions (description_list), rights statements (rights_list), subjects with a scheme (subject_list), geolocations (geolocation_list), contributors (contributor_list), journal identifiers other than ISSN (journal_id_list[]), alternate identifiers of types other than PMID, BIBCODE and URN, the article number (article_number), and the relations in relation_list[] other than the full text PDF.