Relation for occuring emotions

[p.da...@googlemail.com]

Hi there, 

I am using MFOEM in a project of tracking peoples emotions by various sensors. 

A requirement for my database is therefore to connect a occuring emotion from MFOEM to a patient. That patient will most likely be taken from some medical ontology also founding in BFO. 

What I look for is some "best practice" advise for which relation (from what ontology) to use to connect an emotion to a human being.

By investigating why BFO does not deliver any relations in v2.0 I stumbled across Relation Ontology. Maybe some relation taken from there is the solution? Or would you suggest to make up my own predicate? 

Thanks a lot for your help (I hope that group is not dead). 

Kind Regards, 

Patrick Daßler

[Werner Ceusters]

See attached

--
You received this message because you are subscribed to the Google Groups "Emotion Ontology" group.
To unsubscribe from this group and stop receiving emails from it, send an email to emotion-ontology+unsubscribe@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[p.da...@googlemail.com]

Hi, 

thanks, I was aware of that paper. 

Unfortunately it doesn't answer my question. I am looking for a good relation (the middle part of an rdf triple) to connect a human being (my subject) to a emotion (my object). 

Kind regards from Germany, 

Patrick Daßler

See attached

To unsubscribe from this group and stop receiving emails from it, send an email to emotion-ontolo...@googlegroups.com.

[Janna Hastings]

Dear Patrick,

The relation between the person and the emotion is a special kind of participation. The emotion is an occurrent, happening within a given temporal range. The participation is special in several ways, one of which is that it is "private": only one person can participate in an emotion in that way. Emotions are a part of the BFO "history" for the person, which is the sum total of all occurrents taking place during a material entity's life.

We could potentially create an MFOEM-specific relation "has_emotion" as a sub-relation of participates-in in RO (RO_0000056) to link a person to an emotion. Would that help?

Best wishes,

Janna

To unsubscribe from this group and stop receiving emails from it, send an email to emotion-ontology+unsubscribe@googlegroups.com.

[Barry Smith]

The problem is that 'has_emotion' would be ambiguous has between the continuant and occurrent sense. For occurrent emotions we require something like 

person p has_emotion e

e exists_at_temporal_region t

e instance of emotion-type T

BS

On Tue, Mar 21, 2017 at 6:28 AM, Janna Hastings <janna.h...@gmail.com> wrote:

Dear Patrick,

The relation between the person and the emotion is a special kind of participation. The emotion is an occurrent, happening within a given temporal range. The participation is special in several ways, one of which is that it is "private": only one person can participate in an emotion in that way. Emotions are a part of the BFO "history" for the person, which is the sum total of all occurrents taking place during a material entity's life.

We could potentially create an MFOEM-specific relation "has_emotion" as a sub-relation of participates-in in RO (RO_0000056) to link a person to an emotion. Would that help?

Best wishes,

Janna

To unsubscribe from this group and stop receiving emails from it, send an email to emotion-ontology+unsubscribe@googlegroups.com.

For more options, visit https://groups.google.com/d/optout.

--
You received this message because you are subscribed to the Google Groups "Emotion Ontology" group.

To unsubscribe from this group and stop receiving emails from it, send an email to emotion-ontology+unsubscribe@googlegroups.com.

[Janna Hastings]

I was thinking only of emotions which are occurrents. Indeed we separated "emotion process" (most of the emotions e.g. anger) from "emotion disposition" (currently only hate and love). But for the use case being discussed, surely only occurrent emotions are relevant?

In any case, would it not be better for users to distinguish these cases by using different relations rather than complex expressions, so has_occurrent_emotion would becomes a separate relation to has_continuant_emotion? (or, has_emotion_process vs. has_emotion_disposition)

[Patrick Daßler]

Hi there, 

thanks for all those quick responses. I think I am fine with using "Participates in" from RO. 

I expected "has_emotion" to be part of the ontology and was suprised to see that it is not. Please note that I am neither an Ontology- nor an Affective-Science expert. 

As Janna assumed correctly, only occuring emotions are relevant to my use case at the moment. From an non-expert point of view however, implementing an has_occurrent_emotion and has_continuant_emotion seems logical and somewhat straightforward.

Kind Regards from GE, 

Patrick 

--
You received this message because you are subscribed to a topic in the Google Groups "Emotion Ontology" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/emotion-ontology/iJ6q3NWs8MI/unsubscribe.
To unsubscribe from this group and all its topics, send an email to emotion-ontology+unsubscribe@googlegroups.com.

[Stefan Schulz]

Dear all, 

I have always considered good practice in ontology engineering to avoid the proliferation of relations, because this would produce a lot of new primitives and thus reduce precision and affect interoperability. 

Given that BFO2 has a well defined relation "participates in" and the domain ontology has equally well-defined classes Person and Emotion, an additional relation "has emotion" would not provide any benefit for modelling. 

This does not mean that from an application point of view a predicate "has emotion" might be useful. This could be included in a rule like 

a has emotion b =def. a 'instance of' Person, b 'instance of' Emotion, a 'participates in' b

-

Stefan

STEFAN SCHULZ  

(Univ.-Prof. Dr. med.)

Institute for Medical Informatics, Statistics and Documentation
Medical University of Graz

Auenbruggerplatz 2/V, 8036 Graz (Austria)

+43 316 38516939 / +43 699 15096270 / +43 316 38513201 / Skype: stschulz

 

Home: Wiener Str. 5/4, 8020 Graz (Austria)   

[Janna Hastings]

Dear Stefan,

I lean towards agreeing, with the proviso that the label for that relation should be something like "has occurring emotion" rather than just "has emotion" to avoid any ambiguity.

Cheers, Janna

On Mon, Mar 27, 2017 at 11:25 AM, Stefan Schulz <stefan...@medunigraz.at> wrote:

Dear all, 

I have always considered good practice in ontology engineering to avoid the proliferation of relations, because this would produce a lot of new primitives and thus reduce precision and affect interoperability. 

Given that BFO2 has a well defined relation "participates in" and the domain ontology has equally well-defined classes Person and Emotion, an additional relation "has emotion" would not provide any benefit for modelling. 

This does not mean that from an application point of view a predicate "has emotion" might be useful. This could be included in a rule like 

a has emotion b =def. a 'instance of' Person, b 'instance of' Emotion, a 'participates in' b

-

Stefan

[wceu...@gmail.com]

Since you are dealing with continuants there should be time-indexing or you can’t claim compatibility with BFO.

 

a has emotion b at t =def. a 'instance of' Person at t, b 'instance of' Emotion at t, a 'participates in' b at t

 

 

W

 

From: Janna Hastings

Sent: Monday, March 27, 2017 6:31 AM

To: emotion-...@googlegroups.com

Cc: Stefan Schulz

Subject: Re: Relation for occuring emotions

 

Dear Stefan,

I lean towards agreeing, with the proviso that the label for that relation should be something like "has occurring emotion" rather than just "has emotion" to avoid any ambiguity.

Cheers, Janna

On Mon, Mar 27, 2017 at 11:25 AM, Stefan Schulz <stefan...@medunigraz.at> wrote: