Internship topic description
Virtual Observatory / astronomy data sharing / interoperable access to observation data
Table Access Protocol (TAP) (https://www.ivoa.net/documents/TAP/20190927/) is a universal access standard for astronomical relational databases. It is based on an extended, specialized SQL language for astronomy called ADQL (https://www.ivoa.net/documents/ADQL/20180112/PR-ADQL-2.1-20180112.pdf).
The data exposed in a TAP service is described in a schema named TAP_SCHEMA. This description consists of both metadata (name, data type and units) and semantic labels such as a textual description, the semantic class of the quantity described or a pointer to a node of a data model. The TAP_SCHEMA also indicates how tables can be joined. These join rules tell the user how to retrieve complex data elements that are distributed across multiple tables. It is important to note that in a TAP service, metadata is stored in relational tables, i.e. at the same level as the data. Thus, they can be explored with the same tools as for data retrieval. The content of the resource /tables gives access to an XML representation of the TAP_SCHEMA.
Query responses are presented in the form of a single table, which may be denormalized when the query requires joins. They are delivered in an xml table description format called VOTable standardized in the IVOA: (https://www.ivoa.net/documents/VOTable/20191021/REC-VOTable-1.4-20191021.pdf)
The Virtual Observatory (IVOA) (http://ivoa.net) also has a set of standardized data models allowing a logical description of the relationships between various astronomical data. There is currently no standard way to link a table resulting from a TAP query with the IVOA model corresponding to the returned data.
The ModelInstanceIntoVOT project, available in a preliminary version on this git repository (https://github.com/ivoa-std/ModelInstanceInVot#doc-preview) is a proposed standard for mapping model characteristics to one or more tables in VOTable format using an annotation superstructure that allows the different instances and attributes of the model to be found in the table.
The internship includes an exploratory component as well as an implementation component:
- Test various methods or formalism strategies (tables, views, XML or JSON components, ...) to give the TAP service the information needed to map the datasets concerned.
- On-the-fly annotation of ADQL query results: add to an existing TAP server an overlay allowing to insert in the VOTables an XML block including these annotations and allowing the client to relate the data read in the VOTable to the data model sheets.
Translated with www.DeepL.com/Translator (free version)