Re: Add Sweet Ontology to SDWBP 6.2 Expressing Spatial Data

It’s worth remembering that SWEET started as a classification taxonomy — and organization of search terms. It has only really been retroactively labeled an ontology and perhaps that’s unfair.

Josh

> On Apr 4, 2016, at 8:15 AM, Krzysztof Janowicz <janowicz@ucsb.edu> wrote:
> 
> Hi Simon,
> 
> Thanks for bringing this up. I agree that SWEET is a very powerful and useful idea but personally I stopped using it (and I know of many others that did so as well) because of the issues you mentioned as well as many modeling decisions that seemed questionable. This is something that is nearly unavoidable when developing a large structure like SWEET but it makes me feel that in its current stage SWEET should not be part of any best practice. Last time I checked it was also very lightweight, i.e., mostly using only subsumption relations, which limits its applicability.  A reworked version of SWEET would, of course, be great. 
> 
> Best,
> Krzysztof
> 
> 
> On 04/03/2016 10:28 PM, Simon.Cox@csiro.au <mailto:Simon.Cox@csiro.au> wrote:
>> Hi Lewis –
>>  
>> Since you have raised SWEET, it is perhaps worth noting a few ways in which it does not meet the expectations of the linked data community:
>>  
>> (i)                 Every concept (class or property) defined in SWEET is denoted with a http URI, but *these do not resolve* using HTTP. The main way to use SWEET is to download the files, which are available, but huge and take a long time to download;
>> (ii)               The same concept has a *different URI in different versions* of SWEET. In particular the URIs have the SWEET version number in them. The URIs also reflect the factoring between modules, which changes between versions. There is also no tracking back from later versions to older versions, so there is no way to automatically detect semantic equivalence between versions;
>> (iii)             There is almost *no documentation* – no rdfs:label, dc:description, rdfs:comment etc and also no rdfs:seeAlso, skos:closeMatch, rdfs:isDefinedBy etc. So all you have to go by is the name and position in the subsumption hierarchy.
>>  
>> This is a shame, because SWEET is a well thought-out, well-structured resource, but falls short in these few ways. I have raised these issues with Thomas Huang (JPL) who I think is the current maintainer, but haven’t had any impact yet L
>>  
>> Simon
>>  
>> From: lewis john mcgibbney [mailto:lewismc@apache.org <mailto:lewismc@apache.org>] 
>> Sent: Wednesday, 30 March 2016 6:32 PM
>> To: SDW WG Public List <public-sdw-wg@w3.org> <mailto:public-sdw-wg@w3.org>
>> Subject: Add Sweet Ontology to SDWBP 6.2 Expressing Spatial Data
>>  
>> Hi Folks,
>> I would like to propose the addition of the spatially-relevant portions of the Semantic Web for Earth and Environmental Terminology (SWEET) [0] ontology which was authored @JPL.
>> SWEET was developed with the aim of better locating NASA Earth science data with it containing mutual relationships of scientific concepts and their ancillary space, time, and environmental descriptors. There are a number of spatial components which I would be happy to expand upon if required.
>> Thank you for any feedback.
>> Lewis
>> 
>> [0] http://sweet.jpl.nasa.gov <http://sweet.jpl.nasa.gov/>
> 
> -- 
> Krzysztof Janowicz
> 
> Geography Department, University of California, Santa Barbara 
> 4830 Ellison Hall, Santa Barbara, CA 93106-4060 
> 
> Email: jano@geog.ucsb.edu <mailto:jano@geog.ucsb.edu>
> Webpage: http://geog.ucsb.edu/~jano/ <http://geog.ucsb.edu/~jano/>
> Semantic Web Journal: http://www.semantic-web-journal.net <http://www.semantic-web-journal.net/>