dnet-hadoop/dhp-workflows/dhp-aggregation/src/site/markdown/pubmed.md

67 lines
2.4 KiB
Markdown
Raw Normal View History

#Pubmed Mapping
This section describes the mapping implemented for [MEDLINE/PubMed](https://pubmed.ncbi.nlm.nih.gov/).
Collection
---------
The native data is collected from [ftp baseline](https://ftp.ncbi.nlm.nih.gov/pubmed/baseline/) containing XML with
the following [schema](https://www.nlm.nih.gov/bsd/licensee/elements_descriptions.html)
Parsing
-------
The resposible class of parsing is [PMParser](./scaladocs/#eu.dnetlib.dhp.sx.bio.pubmed.PMParser) that generates
an intermediate mapping of PubMed Article defined [here](/apidocs/eu/dnetlib/dhp/sx/bio/pubmed/package-summary.html)
Mapping
-------
The table below describes the mapping from the XML Native to the OAF mapping
| Xpath Source | Oaf Field | Notes |
| ----------- | ----------- | ----------- |
| //PMID | pid | classid = classname = pmid
| | **Instance Mapping** | |
|//PublicationType | InstanceType | If the article contains the typology **Journal Article** then we apply this type else We have to find a terms that match the vocabulary otherwise we discard it
|//PMID | instance/PID | Map the pmid also in the pid in the instance |
| //ArticleId[./@IdType="doi" | instance/alternateIdentifier |classid = classname = doi
|//PMID | instance/URL | prepend to the PMId the base url https://pubmed.ncbi.nlm.nih.gov/
| //PubmedPubDate | instance/Dateofacceptance | apply the function GraphCleaningFunctions.cleanDate before assign it
| FOR ALL INSTANCE | CollectedFrom | datasourceName: *Europe PubMed Central* DatasourceId:
| | **Journal Mapping** | |
|//Journal/PubDate| Journal/Conferencedate | map the date of the Journal
|//Journal/Title| Journal/Name | |
|//Journal/Volume| Journal/Vol | |
|//Journal/ISSN| Journal/issPrinted | |
|//Journal/Issue| Journal/Iss | |
| | **Publication Mapping** | |
| //PubmedPubDate | Dateofacceptance | apply the function GraphCleaningFunctions.cleanDate before assign it
| //Title | title | with qualifier ModelConstants.MAIN_TITLE_QUALIFIER
| //AbstractText | Description ||
|//Language| Language| cleaning vocabulary -> dnet:languages
|//DescriptorName| Subject | classId, className = keyword
| | **Author Mapping** | |
|//Author/LastName| author.Surname| |
|//Author/ForeName| author.Forename| |
|//Author/FullName| author.Forename| Concatenation of forname + lastName if exist |
|FOR ALL AUTHOR | author.rank| sequential number starting from 1|
#TODO
Missing item mapped