<kaz> scribenick: mjkoster

<kaz> scribe: Michael_Koster

prep for f2f

darko: focus on 3 areas

semantic models, TD recipes, semantic processing

darko: review the 3 points, where we are and how to proceed
... 1st point is dsr, but he is not present
... any additional points?
... presents "Linked Data for WoT Applications"
... focus on semantic annotation and discovery
... also could look at iot.schema.org capabilities

mkoster: iot.schema.org is a common pattern for actions, events, properties that can use vocabulary from other ontologies

victor: starting to work on concepts for things and interaction patterns

darko: need to cover the use case for interaction between 2 devices e.g. temperature sensor and air conditioner
... the use case involves replacing a broken sensor with an alternate sensor
... with equivalent capability
... make the model accessible in some repositiory so people can reuse

<victor> http://iot.linkeddata.es/def/core/index-en.html

please review and discuss

darko: semantic processing part
... 2 points, validation and semantic interoperability

victor: trying to use TD ontology to do validation

darko: 2 other ways, using JSON-schema and what dsr is working on using shapes expressions

victor: validation against a set of constraints applied to a Thing Description

danh: semantic processing tools for reasoning
... alignment between SAREF, SSN, SOSA
... trying to use a js reasoner to map vocabularies using alignment rules e.g. subclass
... so discovery using one term will find things exposing all semantically equivalent terms
... will need to develop the mapping rules

darko: different companies may use different semantic models
... this is one way to interoperate between them

danh: hope to be ready with a demonstration at the F2F
... intergrated with scripting API to enable querying with one vocab and discovering things that use another vocab

darko: can we collect examples of events, actions, properties in thing descriptions
... could iot.schema.org be helpful in this

mjkoster: event, action, and property meta model should make it easier to align concepts

TD recipes

darko: TD recipes

darko: recipes enable the orchestration of things
... recipes contain ingredients from TDs with external interactions between them
... thing to thing interactions
... recipes also contain non-functional properties for bindings to locations, etc.
... review of recipe example based on the simple motion sensor-light orchestration
... the recipe specifies how to do discovery as well as the orchestration and binding as separate phases in the workflow

mccool: this is like like IFTTT

darko: more general than IFTTT to allow a wider range of behavior and more powerful

mjkoster: There could be a marketplace for these

darko: next agenda topic?
... roll up and summarize what we just discussed

Danh: create a reference, collect our thoughts in one place

darko: collect and develop it as a reference for the plugfest

maria: explains the Vicinity Core Model linked above
... extending SAREF and SSN ontologies
... also there are SAREF extensions for buildings
... has many devices defined in a cross-domain way
... also defining additional concepts like agents

darko: AOB?
... communicate via email and will set up a common workspace document
... adjourned