W3C home > Mailing lists > Public > semantic-web@w3.org > June 2020

Re: Ontology reasoning - Situation Recognition - Understanding Sensors Data

From: Armin Haller <armin.haller@anu.edu.au>
Date: Tue, 23 Jun 2020 10:15:59 +0000
To: angin scribe <anginscribe3@gmail.com>, Semantic web <semantic-web@w3c.org>
Message-ID: <96F6B596-2C69-4AF2-8521-106FA16E82B2@anu.edu.au>
Hi Angin,

If I understand you correctly you want to determine the context an agent is currently in, based on observed properties of a feature of interest observed by a device? To the best of my knowledge there is no such ontology. What that requires also is more a knowledge graph with an infinite number of correlations between observed properties and the feature of interest they (can) relate to rather than an ontology per se. An example would be the observed speed of a car by a sensor, and that if it is > 0, it means the car is moving, and as such, all passengers within the car are moving too. From an ontological point of view you would need to model the agents occupying a car (foaf:agent, schema:car), the sensors observing the speed (sosa:Sensor, sosa:ObservedProperty), the feature of interest (sosa:FeatureOfInterest), the quantity kind (quantitykind:Speed), the unit of measure (unit:KiloM-PER-HR), and the event/activity that the car is moving (prov:Event/prov:Activity, wot:EventAffordance). You also need to define the rule, where you can use https://www.w3.org/Submission/SWRL/, SPIN https://spinrdf.org/spin.html etc. and that is part of a contextual knowledge graph, rather than part of the ontology.

The namespace prefixes used above are:
foaf: http://xmlns.com/foaf/0.1/

schema: http://schema.org/

sosa: http://www.w3.org/ns/sosa/

quantitykind: http://qudt.org/2.1/vocab/quantitykind

unit: http://qudt.org/vocab/unit/

prov: http://www.w3.org/ns/prov#

wot: https://www.w3.org/2019/wot/td/v1

There is an upcoming Special Issue in the Semantic Web Journal where we discuss which ontologies are available on the Web to model examples such as those, and where there are still gaps in current ontologies. The graph at: https://www.dropbox.com/s/rxsiruuvqr5nlxj/SWJOntologyMainConcepts.pdf?dl=0 is from that editorial and gives an overview of the ontologies and their scope.

Kind regards,

From: angin scribe <anginscribe3@gmail.com>
Date: Tuesday, 23 June 2020 at 6:51 pm
To: Semantic web <semantic-web@w3c.org>
Subject: Ontology reasoning - Situation Recognition - Understanding Sensors Data
Resent-From: <semantic-web@w3.org>
Resent-Date: Tuesday, 23 June 2020 at 6:47 pm

Hello Everyone,

Is there any ontology for reasoning about the situation?

In other words, an ontology for understanding/recognizing the current situation based on the information provided by some sensors, or an ontology that can be used to infer implicit information regarding the current situation based on sensor data.

Could anyone give a pointer to the work in this area?

Many thanks!

Received on Tuesday, 23 June 2020 10:16:19 UTC

This archive was generated by hypermail 2.4.0 : Tuesday, 5 July 2022 08:46:04 UTC