W3C home > Mailing lists > Public > public-gld-wg@w3.org > March 2012

Re: Observations and Measurement - draft

From: Luis Bermudez <lbermudez@opengeospatial.org>
Date: Fri, 9 Mar 2012 09:55:55 -0500
Cc: Benedikt Kämpgen <kaempgen@fzi.de>, "public-gld-wg@w3.org" <public-gld-wg@w3.org>
Message-Id: <5D49CB33-830D-4B58-A5C3-901B1D76FB04@opengeospatial.org>
To: "Gillman, Daniel - BLS" <Gillman.Daniel@bls.gov>
Hi Dan

Good questions.

Observation and Measurement  (O&M)  is the overall container for different concepts related to observations: Procedure, Property, Feature of Interest and Result. I recently learned that there is an O&M ontology [2] developed by CSIRO. Simon Cox is the editor of O&M and he also developed this ontology. I think it is fairly new so I do not think there are a lot of examples out there. However some examples of O&M in XML can be found at the OGC Schemas repositories [1]. This can help clarify what you can do with O&M. O&M also provides details of how the result can be expressed depending on the result type. For example the ontology has subclasses based on the type of observation, such as CategoryObservation, CountObservations and GeometryObservation.

The Result not only contains the values but also the units of measure. Dealing with Units of Measure is at another level and should be treated separately. A units of measure ontology should contain enough information that lets you know how to make conversions from one unit to another, or if two values can be comparable. Similarly, the same applies for the other main  components of O&M, like Feature of Interest or Procedure. They should all have specialize ontologies. For example different ontologies can exist and be plugged in that properly describe procedures for statistical methods, numerical models or sensors.

Within OGC there are two ways to express Units of Measure: 1) via URIs, defined in an ontology or 2) using UCUM [3]. UCUM is very well designed. It is based on base units and nomenclature that can be use to express all the other units. It has also an API. I know ontolog [4] also worked on a units ontology and they looked at UCUM. I'm not sure about the status of the ontolog work but I can ask. I also know that UCUM deals with the dimension issue you mentioned. See specification, Section 3 semantics [5] and example for Healthcare Units [6].


[1] http://schemas.opengis.net/om/2.0/examples/
[2] http://def.seegrid.csiro.au/ontology/isotc211/om
[3] http://unitsofmeasure.org/
[4] http://ontolog.cim3.net/cgi-bin/wiki.pl?UoM_Ontology_Standard
[5] http://aurora.regenstrief.org/~ucum/ucum.html
[5] http://www.hl7.de/download/documents/ucum/ucumdata.html

Best Regards,

Luis
-------------
Luis Bermudez
lbermudez@opengeospatial.org  
+1 (301) 760-7323 

The OGC: Making Location Count...
http://www.opengeospatial.org/contact




On Mar 8, 2012, at 5:09 PM, Gillman, Daniel - BLS wrote:

> Luis,
> 
> In this Observation and Measurement scheme, is there a way to record equivalent units of measure?  Suppose I am measuring the speed of a car using a radar gun on some road at a dangerous curve where people usually drive too fast, and the radar gun reports the speed in miles per hour.  Then, suppose someone else repeats the experiment using the same radar gun, but the machine is set to report kilometers per hour instead.  How do we know that all data reported in mph can be converted to data in kph, and vice-versa?
> 
> The ISO/IEC 11179 standard (http://standards.iso.org/ittf/PubliclyAvailableStandards/index.html) introduced the notion of dimensionality, the name coming from the dimensional analysis that scientists use to convert units of measure.  However, this takes a wider view to include units of measure that are not a part of the scientific set, and the idea doesn't depend on the faux algebra of cancelling the same units in numerator and denominator.
> 
> Dimensionality depends chiefly on the set of operations, very similar to datatypes, that data in a unit of measure are allowed.  Take temperatures measured in degrees Celsius.  One can subtract them, take averages and similar statistics, but scalar multiplication makes no sense.  For instance, twice 10C is not 20C.  Therefore, data in degrees Fahrenheit are convertible to Celsius, since they share the same operations.  Interestingly, under this definition, Kelvin is not an equivalent measuring system, and that is because scalar multiplication makes sense in Kelvin.
> 
> Yours,
> Dan
> 
> 
> Dan Gillman
> Bureau of Labor Statistics
> Office of Survey Methods Research
> 2 Massachusetts Ave, NE
> Washington, DC 20212 USA
> Tel     +1.202.691.7523
> FAX    +1.202.691.7426
> Email  Gillman.Daniel@BLS.Gov
> ----------------------------------------- 
> "He has all the virtues I dislike
> and none of the vices I admire."
> - Winston Churchill
> ------------------------------------------
> 
> 
> 
> -----Original Message-----
> From: Luis Bermudez [mailto:lbermudez@opengeospatial.org] 
> Sent: Thursday, March 01, 2012 11:10 AM
> To: Benedikt Kämpgen
> Cc: public-gld-wg@w3.org
> Subject: Observations and Measurement - draft
> 
> All,
> 
> Comments about Observations and Measurement related to:
> 
> - Section: Publishing Statistical Data
> - http://www.w3.org/2011/gld/track/issues/32
> - @@TODO: Are there any statements about compatibility and interoperability between O&M and Data Cube that can be made to give guidance to such organizations?
> 
> Suggest to use Observations and Measurements (OM) model.
> 
> An Observation is an event whose result is an estimate of the value of some property of a feature of interest, obtained using a specified procedure.
> 
> An Observation has essentially the following parts:
> 
> Procedure: e.g. person, sensor, observatory, model, macro.
> Property: e.g. temperature, color, weight.
> Feature of Interest: e.g. body, world, Washington DC, banana, specimen 123, Monterey Bay.
> Result: The actual value. Includes categories and quantities results, time series, multimedia (video, audio), one single value, a file, etc.
> 
> So in the figure: "Modeling quantity, measurement, observation" we can do the following:
> 
> Add Procedure, Result, Feature of Interest and Property Model Person as a type of Procedure Remove Measurement Category and Quantity are subclass of result.
> Phenomenon Type is same as Property
> 
> 
> Luis
> -------------
> Luis Bermudez
> lbermudez@opengeospatial.org  
> +1 (301) 760-7323
> 
> The OGC: Making Location Count...
> http://www.opengeospatial.org/contact
> 
> 
> 
> 
> On Feb 22, 2012, at 11:38 AM, Benedikt Kämpgen wrote:
> 
>> Hello,
>> 
>> We have further refined the Data Cube use cases and copied the content 
>> from the wiki into an Editor's Draft for "Use Cases and Requirements 
>> for the Data Cube Vocabulary" [1].
>> 
>> Feedback of any kind would be well appreciated.
>> 
>> Best,
>> 
>> Benedikt
>> 
>> [1] 
>> <http://dvcs.w3.org/hg/gld/raw-file/default/data-cube-ucr/index.html>
>> 
>> 
>> 
>> --
>> AIFB, Karlsruhe Institute of Technology (KIT)
>> Phone: +49 721 608-47946
>> Email: benedikt.kaempgen@kit.edu
>> Web: http://www.aifb.kit.edu/web/Hauptseite/en
>> 
>> 
>> 
>> 
>> 
> 
> 
> 
> 
Received on Friday, 9 March 2012 14:56:22 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 9 December 2014 23:03:31 UTC