[Esip-soil-informatics] FW: Followup questions from 5 May GLOSIS presentation
Raul Palma
rpalma at man.poznan.pl
Thu May 13 16:08:04 EDT 2021
Hi Gary,
Indeed, the ontology was derived from the GLOsis UML model; however, as described in the presentation it was not a direct one to one transformation of UML into OWL. The most important part of the ontology generation was the alignment with the SOSA/SSN model and approach (that is based and extends the SSO pattern).
The final ontology was just the initial step. Right now, for example, colleagues from ISRIC are extending the ontology to include the procedures associated to the different types of soil observations. Similarly, we could on the one hand, extend the ontology to include additional, but relevant, concepts coming from other models (currently missing in the ontology), and we could make alignments between the ontology and other relevant models (if any), e.g., creating equivalentClass, equivalentPropertiy or taxonomic axioms.
In any case, I think a first important step is adding textual (human-readable) descriptions to the ontology terms as much as possible. But this will require a good support from domain experts.
Best Regards,
Raul
From: Esip-soil-informatics <esip-soil-informatics-bounces at lists.esipfed.org> on behalf of Gary Berg-Cross via Esip-soil-informatics <esip-soil-informatics at lists.esipfed.org>
Reply to: Gary Berg-Cross <gbergcross at gmail.com>
Date: Thursday, 13 May 2021 at 18:32
To: Kathi Schleidt <kathi at datacove.eu>
Cc: "ESIP, soil info" <esip-soil-informatics at lists.esipfed.org>
Subject: Re: [Esip-soil-informatics] FW: Followup questions from 5 May GLOSIS presentation
Kathi et al,
A few more thoughts. I hope they are useful.
On your point:
>Raul's task was not to design a soil ontology, but to transform the model for the Global Soil Information System GloSIS towards a semantic representation >thereof.
Moving the data representation to the richer OWL form but in alignment with the original DB is useful, but I would note from Raul's presentation this which does use the concept of an ontology:
"The presentation will provide an overview of the process for transforming the Glosis UML data model into an OWL ontology."
Point (or question) #1 raised the idea that one might refine, harmonize and add to this initial ontology over time with additional, but relevant concepts. These might come from and align with other soil DBs. There may be some definitional differences to resolve if this is attempted and the ontology can serve as an integration point and a conceptual basis for such growth.
This does also touch on point #2,
It is true that no one ontology may be suitable. There may be several but to avoid silos the possibility term and axiom reuse
should be considered. Glosis UML may not be the only basis for common and useful concepts in at least one larger ontology.
To enable this some common soil ontology design patterns (SSO is one such) might be considered. Unlike SSO this would be more soil-centric.
Gary Berg-Cross
Potomac, MD
240-426-0770
On Tue, May 11, 2021 at 1:01 PM Kathi Schleidt via Esip-soil-informatics <esip-soil-informatics at lists.esipfed.org> wrote:
Dear all,
a bit of background from the underlying model that Raul has so nicely transformed to a more semantic representation - please bear in mind that Raul's task was not to design a soil ontology, but to transform the model for the Global Soil Information System GloSIS towards a semantic representation thereof. For details of that data model, please go back to the talk that Tomas Reznik and myself held for this group at the April 7th meeting.
To point 1 and the definition of bulk density (pretty sure you'll encounter similar issues in a lot of the concepts for with data is foreseen) - all the requirements were extracted from the FAO Guidelines on Soil Description as well as the available drafts towards GloSIS - thus the semantics of these concepts should be aligned with those sources, ideally with domain experts, which we data geeks are definitely NOT (I just understand the structure of what you do, not the details ;) )
To point 2 representation of the soil domain - while I think the world is still looking for ONE ontology to support all requirements, I've become wary of this goal, as ontologies are strongly guided in their perspective on the world by their use cases - as GloSIS was designed to collect, harmonize, merge all available data on soil, it took a measurement-data-centric perspective, leading to the statement of:
"So, the domain knowledge is there but is not modelled as subclasses of soil concepts, but as types of observations of soil concepts"
I totally get the requirement for a more user-centric perspective on soil, but believe this should be a complementary (but interlinked) ontology representing user concepts and thus suitable for answering questions such as posed in your example of "is this soil compact enough, healthy enough or too acidic to support this type (X) of farming?" - if suitably interlinked, the user-centric system could then transpose the question of if "farming type X can be supported" to the measurement-data-centric perspective provided by GloSIS (accessing information on compactness, healthiness or acidity), aggregate this measurement data, put it in context, and return an answer.
Does this help?
:)
Kathi
On 11.05.2021 18:05, Gary Berg-Cross via Esip-soil-informatics wrote:
Raul et al,
Thank you for the follow up response. I hope that some of the issues and understandings will
be helpful down the road.
I can understand the idea that a definition is implied in the collection axioms such as
you have for BulkDensityWholeSoil using the axioms
is a subclass of sosa:Observation
hasFeatureOfInterest a Layer or a Horizon
the observedProperty is bulkDensityWholeSoilProperty
and the expected values are numeric (float) measured in KiloGM-PER-DeciM3
I do see some issue with the variance here to what is in typical definitions of Bulk Density
What it is: Bulk density is an indicator of soil compaction. It is calculated as the dry weight of soil divided by its volume. This volume includes the volume of soil particles and the volume of pores among soil particles. Bulk density is typically expressed in g/cm3.
Adding definitions like this is helpful to human understanding.
This also suggests, of course, that definitions promote the idea some additional axioms including the relation to soil compaction (soil has part compaction process) and a calculation process.
EnvO, for example, has this definition for: compaction process which might be used as context:
def: "A physical process during which atoms, molecules, or other constituents of a material entity are forced closer together."
On the 2nd issue of how the soil domain is represented, I see that, as you say:
"So, the domain knowledge is there but is not modelled as subclasses of soil concepts, but as types of observations of soil concepts"
I wonder if people would agree that there is an understood soil domain which includes features (of interest) and that based on this integrated understanding observations are made to investigate the domain. Particular observations may be of interest for different types of soil and for purposes such as - "is this soil compact enough, healthy enough or too acidic to support this type (X) of farming?"
Gary Berg-Cross
Potomac, MD
240-426-0770
On Fri, May 7, 2021 at 12:35 PM Todd-Brown, Kathe via Esip-soil-informatics <esip-soil-informatics at lists.esipfed.org> wrote:
From: Raul Palma <rpalma at man.poznan.pl>
Date: Friday, May 7, 2021 at 5:13 AM
To: "Egmond, Fenny van" <fenny.vanegmond at wur.nl>, Luís de Sousa <luis.desousa at isric.org>, Tomáš Řezník <tomas.reznik at sci.muni.cz>, Kathi Schleidt <kathi at datacove.eu>, "Bulens, Jandirk" <jandirk.bulens at wur.nl>, Bogusz Janiak <bogusz.janiak at gmail.com>
Cc: "Todd-Brown, Kathe" <kathe.toddbrown at essie.ufl.edu>
Subject: Re: [Esip-soil-informatics] Followup questions from 5 May GLOSIS presentation
Hi,
Sure, from my side I can say a few words.
Regarding question 1.
Indeed, the concepts are not having a textual description, apart from the codelists: https://github.com/rapw3k/glosis/blob/master/glosis_cl_v1.0.0.ttl
The definition is at the owl level, so for example,
BulkDensityWholeSoil we can see that
is a subclass of sosa:Observation
hasFeatureOfInterest a Layer or a Horizon
the observedProperty is bulkDensityWholeSoilProperty
and the expected values are numeric (float) measured in KiloGM-PER-DeciM3
Nevertheless, I agree having also textual descriptions would be a good addition.
This would require, though, to go through documentation and excels of the GLOSIS, but should be done.
Regarding question 2.
I think the view on this is a bit reverse.
The main concept taxonomy of the soil domain is derived from the ISO28258, as you can see in the picture attached.
Then all the observation types that we have defined in GloSIS are connected to those soil concepts (via the hasFeatureOfInterest property).
So, the domain knowledge is there but is not modelled as subclasses of soil concepts, but as types of observations of soil concepts.
To make it more explicit, we could make a superclass for all those observations, e.g., , something like “SoilObservation” (as a subclass of sosa:Observation) and then make all our observations subclasses of that superclass (instead of being subclass of sosa:Observation). This way it could be clear that these are observations of soil concepts, but in terms of explicit semantics is not really adding much.
Perhaps others here have some opinions in this regard?
Cheers,
Raul
From: "Egmond, Fenny van" <fenny.vanegmond at wur.nl>
Date: Friday, 7 May 2021 at 10:38
To: Raul Palma <rpalma at man.poznan.pl>, Luís de Sousa <luis.desousa at isric.org>, Tomáš Řezník <tomas.reznik at sci.muni.cz>, Kathi Schleidt <kathi at datacove.eu>, "Bulens, Jandirk" <jandirk.bulens at wur.nl>
Cc: "Todd-Brown, Kathe" <kathe.toddbrown at essie.ufl.edu>
Subject: FW: [Esip-soil-informatics] Followup questions from 5 May GLOSIS presentation
Hi,
I don’t know if you receive the emails on the emailstring below, so just to be sure I forward to you. Would one of you like to answer to Gary through the mailing list?
Thanks!
Fenny
-----------------------------------------------------------------------------------------------------------------------------------------------------
Fenny van Egmond
Wageningen Environmental Research (and ISRIC – World Soil Information)
Postbus 47, 6700 AA, Wageningen
Wageningen Campus, Gebouw 101 (Gaia), Room G.C.010 (mon, tu, fr) and Room C.111 (wed, th, fr)
Droevendaalsesteeg 3, 6708 PB, Wageningen
T. +31(0)317 480406
E. fenny.vanegmond at wur.nl
www.wur.nl/environmental-research
From: Esip-soil-informatics <esip-soil-informatics-bounces at lists.esipfed.org> On Behalf Of Gary Berg-Cross via Esip-soil-informatics
Sent: Thursday, May 06, 2021 19:38
To: Todd-Brown, Kathe <kathe.toddbrown at essie.ufl.edu>
Cc: ESIP, soil info <esip-soil-informatics at lists.esipfed.org>
Subject: Re: [Esip-soil-informatics] Followup questions from 5 May GLOSIS presentation
Raul,Et al.
I have two followup questions from the GLOSIS ontology presentation that there wasn't time
to cover during our hour session.
1. Are there definitions in the GLOSIS ontology?
When I look at a class in the ontology like " BulkDensityWholeSoil" I don't see definitions.
I used the link:
http://w3id.org/glosis/model/layerhorizon#BulkDensityWholeSoil
And I can see a text version of the Owl file.
For BulkDensityWholeSoil it is this segment info:
glosis_lh:BulkDensityWholeSoil a
owl:Class ;
rdfs:subClassOf sosa:Observation ;
rdfs:subClassOf [ a owl:Restriction ; owl:onProperty sosa:hasFeatureOfInterest ; owl:allValuesFrom [owl:unionOf (glosis_lh:GL_Layer glosis_lh:GL_Horizon) ] ] ;
rdfs:subClassOf [ a owl:Restriction ; owl:onProperty sosa:hasResult ; owl:allValuesFrom glosis_lh:BulkDensityWholeSoilValue ] ;
rdfs:subClassOf [ a owl:Restriction ; owl:onProperty sosa:observedProperty ; owl:hasValue glosis_lh:bulkDensityWholeSoilProperty ] .
glosis_lh:bulkDensityWholeSoilProperty a sosa:ObservableProperty , qudt:Quantitykind ;
rdfs:label "bulkDensityWholeSoilProperty"@en .
glosis_lh:BulkDensityWholeSoilValue a owl:Class ;
rdfs:subClassOf qudt:QuantityValue ;
rdfs:subClassOf [ a owl:Restriction ; owl:onProperty qudt:numericValue ; owl:allValuesFrom xsd:float ] ;
rdfs:subClassOf [ a owl:Restriction ; owl:onProperty qudt:unit ; owl:hasValue unit:KiloGM-PER-DeciM3] .
So no commented definition for this concept. In general I don't see definitions, but would believe you have them, they are useful for understanding the model of the domain. What am I missing?
2. I'd like to know more about thoughts on the "domain" taxonomy and structure.
I see ample organization using concepts like sites and things that Observations out of SOSA and SSN.
These are bottom line useful for soil data by code etc.
But there seems to
be little about the soil domain itself (like sub-classes).
Am I missing something or is that an issue?
I've attached a portion of the WebVowl view of Glosis which might make clearer
how I understand the domain organization and absence of a Soil concept as an organizer.
This would be useful for a real use case and for harmonizing
across efforts.
Perhaps you have plans here.
Thanks,
Gary Berg-Cross
Ontolog Forum
Potomac, MD
240-426-0770
On Wed, May 5, 2021 at 3:31 PM Todd-Brown, Kathe via Esip-soil-informatics <esip-soil-informatics at lists.esipfed.org> wrote:
Thank you Raul for a fantastic overview of the GloSIS ontology. If you missed it you can find the recording here: https://youtu.be/q79p_oA5rtU and Anne D took fantastic notes in the logistics document.
Brief housekeeping: We now have ONE zoom room for both meetings and the easy and convenient calendar generating links might not auto update. Sorry that this information keeps changing, please double check your calendar to make sure it matches the information on the google document https://docs.google.com/document/d/1_SIm_3e9xpni_64zdAU1DZrFEd8YcIerbsihDJq4qwY/edit?usp=sharing
We will be meeting next on 19 May at 2300UTC. If no one comes forward with a different suggestion, on the agenda will be pulling together a soil semantic resource list for the ESIP Summer meeting (and ourselves!).
We are looking for other presentations! If you know of anyone working with soil data who would be interested in presenting please send them my way!
-Kathe
---------
Katherine Todd-Brown, PhD
Assistant Professor
Environmental Engineering Sciences
Engineering School of Sustainable Infrastructure & Environment
Herbert Wertheim College of Engineering
University of Florida
Office: Rm B006, Phelps Lab, UF main campus
Twitter: @KatheMathBio
Mail: Dr Kathe Todd-Brown
Center for Wetlands
PO Box 116350
Gainesville, FL 32611
_______________________________________________
Esip-soil-informatics mailing list
To start a new topic: Esip-soil-informatics at lists.esipfed.org
To unsubscribe and manage prefs: https://lists.esipfed.org/mailman/listinfo/esip-soil-informatics
_______________________________________________
Esip-soil-informatics mailing list
To start a new topic: Esip-soil-informatics at lists.esipfed.org
To unsubscribe and manage prefs: https://lists.esipfed.org/mailman/listinfo/esip-soil-informatics
_______________________________________________
Esip-soil-informatics mailing list
To start a new topic: Esip-soil-informatics at lists.esipfed.org
To unsubscribe and manage prefs: https://lists.esipfed.org/mailman/listinfo/esip-soil-informatics
--
_________________________________________________________________________
Katharina Schleidt
Tel: +43 (1) 89 234 26
Mobile: +43 (650) 89 234 26
Skype: Kathi Schleidt
Kathi at DataCove.eu
www.DataCove.eu
In the twenty-first century censorship works by flooding people with irrelevant information.
-- Homo Deus, Yuval Noah Harari
Do what you can, when you can, because you can - anonymous, paraphrasing Theodore Roosevelt
Please note that the fact that you have received this email implies that your mail address is stored on my system's address book. If this bothers you, please get in touch, and I will delete your information.
_______________________________________________
Esip-soil-informatics mailing list
To start a new topic: Esip-soil-informatics at lists.esipfed.org
To unsubscribe and manage prefs: https://lists.esipfed.org/mailman/listinfo/esip-soil-informatics
_______________________________________________ Esip-soil-informatics mailing list To start a new topic: Esip-soil-informatics at lists.esipfed.org To unsubscribe and manage prefs: https://lists.esipfed.org/mailman/listinfo/esip-soil-informatics
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.esipfed.org/pipermail/esip-soil-informatics/attachments/20210513/a43f6c45/attachment-0001.htm>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: image001.jpg
Type: image/jpeg
Size: 3321 bytes
Desc: not available
URL: <http://lists.esipfed.org/pipermail/esip-soil-informatics/attachments/20210513/a43f6c45/attachment-0001.jpg>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: image002.png
Type: image/png
Size: 81328 bytes
Desc: not available
URL: <http://lists.esipfed.org/pipermail/esip-soil-informatics/attachments/20210513/a43f6c45/attachment-0001.png>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: smime.p7s
Type: application/pkcs7-signature
Size: 5410 bytes
Desc: not available
URL: <http://lists.esipfed.org/pipermail/esip-soil-informatics/attachments/20210513/a43f6c45/attachment-0001.p7s>
More information about the Esip-soil-informatics
mailing list