INTER-IoT Mapping and Translation

<inserted> Katya's slides: https://lists.w3.org/Archives/Public/www-archive/2018Apr/att-0002/INTER-IoT_Semantic_Interoperability_3-04-2018.pdf

katya: 1. defined mapping def format 2. created tools
... Translating only the messages sent -- not the entire ont.
... Alignment-based translation -- rewriting the RDF graph. We assume a central ont.
... Two alignments normally done: to and from the central ont.
... The central ont we assume is modular to be extendable to the env where it is used.
... We can include exchanges modules for dedicated environments.
... E.g., we could have modules for spacial or medical domains.
... The alignment format we designed is based on the alignment API level 2.
... Very expressive: transformation, external functions and many other functions.
... And a tool can use it to do the translation.
... Alignment format is IPSM-AF.
... We're matching parts of the RDF graph and can list the parts that were not translated.

(slide: "IPSM-AF -- header")

katya: the header includes ont info, title, version, etc.
... Then the logic is in the steps and mappings.
... Each step is consecutive. with order defined in the AF.

eric: Can the mapping for one cell reference the mapping for another?

katya: No, each step works on the output of the previous.

(slide: "Alignment Cell")

katya: entity1 specifies the source pattern, entity2 is the target pattern.
... entity2 specifies how the graph should be rewritten.
... Can call SPARQL functions, or even an ext function.
... Variables from source can be used in the target pattern.

eric: For ext functions, can they generate new rows (like in a SPARQL CONSTRUCT)?
... In a SPARQL CONSTRUCT, the WHERE clause gives you a certain number of rows, which are then instantiated into the CONSTRUCT pattern. Some SPARQL extensions can expand the number of solution rows.

katya: Cannot do that at the moment. Not generating additional rows.

<scribe> ACTION: eric to show example of generating new rows and why it is sometimes needed

(slide "Alignment Excerpt")

katya: This slide shows example translation.
... semantics are the same, but changing the ont and presentation.

eric: Stan Huff calls it "isosemantic"

katya: The XML uses some of the RDF terms, but it is not RDF. It has variables.

david: file is parsable as RDF/XML?

katya: Yes, but the entity1 and entity2 parts then need to be sub-parsed (micro-parsed).
... in order to fully understand it.
... Joao and I also started looking at the FHIR ont for use.
... We are using two platforms, one for data from wearable sensors, which uses JSON.
... The other platform is the universal platform, which supports semantics, and has its own ont defined.
... Recently thought of using the universal ont and map them onto FHIR ont, and see how those two can be translated.

Comments from Joao in the hangout chat window:

[[

Yes, unfortunatelly today I can't speak. But on the next meeting I think we could discuss about the idea of creating an extension of GOIoTP (INTER-IoT core ontology) with FHIR RDF, focusing on ECG data representation

which seems to be done through fhir:Observation.component.valueSampledData according to this example: https://www.hl7.org/fhir/observation-example-sample-data.json.html

UniversAAL IoT platform also allows to include your own ontology. But my tests with this platform aren't going well..

]]

eric: The RDF data validation working group has a split of people who want to do SHACL and people who want to do ShEx. SHACL is based on SPARQL CONSTRUCT; ShEx was green-field designed for validation. And an extension ShExMap allows translation to be done.

Next steps

<scribe> ACTION: Eric to send out prep materials about ShEx / ShExMap

<scribe> ACTION: David to ask Glenna if she can join the call on May 1 to discuss FHIR translation using shex

AGREED: Next discussion will be May 1, Eric presenting shex & shexmap

<inserted> (Eric will be away the next 2 weeks)

ADJOURNED

This is scribe.perl Revision: 1.152  of Date: 2017/02/06 11:04:15  
Check for newer version at http://dev.w3.org/cvsweb/~checkout~/2002/scribe/

Guessing input format: Irssi_ISO8601_Log_Text_Format (score 1.00)

Succeeded: s/João MoreiraYes/Yes/
Succeeded: s/ 9:40 AMJoão Moreira//
Succeeded: s/9:41 AMJoão Moreira//
FAILED: i|Katya's slides: katya: 1. defined mapping|
Succeeded: s/i|Katya's slides: katya: 1. defined mapping|//
Succeeded: i|katya: 1. defined mapping|Katya's slides: https://lists.w3.org/Archives/Public/www-archive/2018Apr/att-0002/INTER-IoT_Semantic_Interoperability_3-04-2018.pdf
Succeeded: i/ADJOURNED/(Eric will be away the next 2 weeks)
Present: Brian_(Kaiser HL7) David_Booth EricP Katya Joao Pawel
No ScribeNick specified.  Guessing ScribeNick: dbooth
Inferring Scribes: dbooth

WARNING: No date found!  Assuming today.  (Hint: Specify
the W3C IRC log URL, and the date will be determined from that.)
Or specify the date like this:
<dbooth> Date: 12 Sep 2002

People with action items: david eric

WARNING: IRC log location not specified!  (You can ignore this 
warning if you do not want the generated minutes to contain 
a link to the original IRC log.)