Wikidata:WikiProject Built heritage/Data sources
|
|
Sources to Import Data From edit
The table below contains an overview over the (official) data sources to import data about built heritage from (listed by country). Try to be as comprehensive as possible when listing potential data sources of a given country.
| Country | Publisher | Database | Coverage | Wikidata item | Mapping information | Rights | Contact | Comments | Ingestion Status |
|---|---|---|---|---|---|---|---|---|---|
| CH | Swiss Federal Office of Civil Protection | KGS Inventar (Inventory of the Protection of Cultural Property in Switzerland) | Cultural property of national (A lists) and regional (B lists) significance in Switzerland. Contains both built heritage and heritage collections. | Swiss Inventory of Cultural Property of National and Regional Significance (Q869941) | CC-0 | Maintainer of the database: kgs-gis@babs.admin.ch
Data ingestion on Wikidata: User:Beat Estermann |
In the future, the A lists (cultural property of national significance) will be published on the opendata.swiss website. | Completed | |
| CH | Canton of Zurich, Monuments Protection Service | Inventar der Denkmalschutzobjekte (Canton of Zurich) | Historical monuments in the Canton of Zurich that have been inventoried at a supra-municipal level | Inventory of historical monuments in the canton of Zurich (Q28971560) | opendata.swiss "Open use" (equivalent to CC-0) | Maintainer of the database: are.denkmalpflege@bd.zh.ch | - | ||
| CH | City of Zurich, Monuments Protection Service | Inventar der kunst- und kulturhistorischen Schutzobjekte der Stadt Zürich | Historical monuments in the City of Zurich (buildings, fountains, bridges, monuments) | historical monuments in the city of Zurich, Switzerland (Q30237745) | CC-0 | Maintainer of the database: martina.jenzer@zuerich.ch | - | ||
| CH | Canton of Aargau | Denkmalschutzinventar (ScopeArchiv) | Historical monuments in the Canton of Aargau | Data cannot easily be downloaded. | Not planned until it is available as open data (technical accessibility); available through the Monuments Database. | ||||
| CH | Canton of Appenzell Ausserrhoden, Denkmalpflege | Listen mit den Schutzobjekten pro Gemeinde (one PDF file per municipality) | Historical monuments in the Canton of Appenzell Ausserrhoden | Data cannot easily be extracted (one PDF file per municipality) | Not planned until it is available as open data (technical accessibility); available through the Monuments Database. | ||||
| CH | Canton of Lucerne, Denkmalpflege | Kantonales Denkmalverzeichnis (one PDF list) | Historical monuments in the Canton of Lucerne | Data cannot easily be extracted (PDF file) | Not planned until it is available as open data (technical accessibility); available through the Monuments Database. | ||||
| CH | Canton of Schwyz, Amt für Kultur | Denkmallisten (one list per municipality in a database) | Historical monuments in the Canton of Schwyz | Data cannot easily be extracted (online database) | Not planned until it is available as open data (technical accessibility); available through the Monuments Database. | ||||
| CH | Canton of Zug, Amt für Denkmalpflege und Archäologie | Inventar und Verzeichnis der Denkmäler (PDF lists) | Historical monuments in the Canton of Zug | Data cannot easily be extracted (two PDF files per municipality) | Not planned until it is available as open data (technical accessibility); available through the Monuments Database. | ||||
| CH | Wiki Loves Monuments / Wikipedia | Monuments Database | German (and partly French) labels of the objects contained in various official databases. Additional historical monuments from cantonal and municipal sources. | ch_(de), ch_(fr), ch_(it), | The data in the Monuments Database has been harvested from monuments lists on German, French, and Italian Wikipedia. With regard to official data sources, they should be used on a subsidiary basis. As the labels of the objects often have been edited by Wikipedians, it would make sense to import the labels from the Monuments Database and to include the original labels from the official databases in a separate statement.
The Monuments Database also contains data about cultural properties of cantonal or regional significance from some cantons where this data has not been made in a technically easily accessible manner. |
Planned |
Describe the Data Sources on Wikidata edit
Databases suitable for data ingestion into Wikidata should be described on Wikidata itself. For each database, an item needs to be created on Wikidata, so that the database can be cited as a source when ingesting data into Wikidata. Refer to Help:Sources for information about how to use sources in Wikidata. When it comes to importing data about built heritage, the types of data sources most commonly found are:
- Online databases: In the case of an online database, a Wikidata property should be created that corresponds to a unique identifier used to refer to items in the database. For the database itself, a Wikidata item should be created. Refer to Help:Sources#Databases for further guidance. Example: <insert an example of an online database here.>.Note: In fall 2016, it was impossible to properly reference online databases using the Quick Statements Tool, a tool commonly used to batch ingest data into Wikidata. Since December 2016, a new version of the tool is under development. <Check the next release of the tool and update this note if necessary.>
- Database dumps / database exports: In the case of a database dump or a database export, the source file is typically available in a spreadsheet format (e.g. CSV or Excel) or in a hierarchical format (e.g. XML). In this case, a Wikidata item should be created for the database itself (example: Swiss GLAM Inventory (Q26933296)) and for the specific export file (example: Swiss GLAM Inventory, 16 September 2016 (Q27477970)).
- Simple web pages: In some cases you may find lists of built heritage on simple web pages. If the source is a simple website, it does not need to be described as a separate Wikidata item. Refer to Help:Sources#Web page for further information about how to source statements to simple web pages.
Unique Identifiers edit
Before ingesting data into Wikidata, we usually want to make sure that our source database contains a unique identifier which can later be used to match the data on Wikidata with the data in the source database. This is particularly useful in the case of future updates to the source database. There are two commonly used approaches to ingesting such unique identifiers into Wikidata:
- The use of a single-source identifier: In this case, a particular Wikidata property is created that corresponds to a unique identifier used in a given database. Example: Elvis Presley (Q303) has a Library of Congress authority ID (P244) : "n78079487". Typically, identifier properties have their corresponding Wikidata entity, in the case of our example: Library of Congress Authorities (Q13219454).
- The use of a multi-source identifier: In this case, a particular Wikidata property and a Wikidata item are used that are generic for a particular domain; multi-source identifiers should be unique in combination with a qualifier. Example: The item Judith Holding the Head of Holofernes (Q17319619) has an inventory number (P217) : "SK-A-1" that is further qualified by collection (P195) : Rijksmuseum (Q190804). In this case, the collection and the inventory number combined form a unique identifier (the inventory number on its own does not have to be unique across collections). The corresponding Wikidata entity in this example is: accession number (Q1417099).
With regard to the data sources concerning built heritage, it is recommended to use single-source identifiers for well-established databases. In this case, you will have to:
- Create a Wikidata entity for the identifier (if it does not already exist).
- Propose the creation of a corresponding property. Note: creating new properties requires community approval that may take several weeks.
If you have compiled an inventory of built heritage in your country by drawing on various sources, the property Wiki Loves Monuments ID (P2186) may be used to add unique identifiers to Wikidata. Before doing so, you should coordinate with the international WLM team that has started using the Wiki Loves Monuments ID as unique identifier; it is important that these identifiers remain unique.
Mapping Between the Data Structure in the Source Files and the Data Structure on Wikidata edit
Another important step in view of the ingestion of data from the source databases into Wikidata is the mapping between the data structure in the source files and the data structure on Wikidata. For this purpose, you should create a sub-page of this page that is specific to your country (unless one already exists). Example: Mapping information for data sources covering Swiss heritage institutions. Replace the example by one relating to built heritage!
The data needs to be mapped at two levels:
- Properties: For each property contained in the source data file, a corresponding property needs to be identified (or newly created) on Wikidata. If the source data is in table format, the column headers usually represent the properties, while each row typically represents one item. A list of properties commonly used in relation to built heritage can be found in the Data Structure section.
- Classes / controlled vocabularies: In some cases, the values of the properties may be simple strings, as for example in a physical address ("Main Street 1"); in this case, no further mapping is needed. In other cases, the values are controlled vocabularies represented on Wikidata by specific classes or entities. In this case, the values in the source data file need to be mapped to those specific classes or entities. See the Mapping Information for the Swiss GLAM Inventory for a series of examples. An overview of controlled vocabularies commonly used in relation to built heritage can be found in the Data Structure section; an overview of the typology of built heritage currently in use on Wikidata can be found in the Typology section. Depending on the dataset and the country, it may be necessary to complement these controlled vocabularies by creating new items on Wikidata before the ingestion of the data can begin.