looking for a good owl pattern

classic Classic list List threaded Threaded
13 messages Options
Reply | Threaded
Open this post in threaded view
|

looking for a good owl pattern

AlexJ
can anybody advise a good OWL pattern or at least point me at a right place:) for the following problem:
there is a need to integrate some educational resources, lets call them Articles under different views like by Knowledge Areas, Alphabetical etc. So, I created the Classes as follows:

Article
  DeepProfessionalArticles
  SurfaceEducationalArticles

Knowledge Area
  Thermodynamics
     Boiling
     etc
  Heat and Mass Transfer
     Heat Exchangers
     etc

Alphabetical View
  A
  B
  C

Now, I have to add individuals for Article, one by one, and then connect them to one or more Knowledge Areas. The most desirable way, from the first glance, would be to create a property hasKnowledgeArea and hasAlphabeticalView and then use it to connect ArticleIndividual to a particular Area. Right now I see only two way:
- create an individual for each Knowledge Area and then connect article individual to KA individual
- create a tree of defined abstract classes as subclasses of Article which will actually duplicate Knowledge Area tree so idividulas will be placed under right class by reasoner

Both of them look like unsatisfied for me... I just wonder if someone was faced with the same problem and are there any patterns for that task? Or just point me to a right direction... The solution looks like to be on the surface and I am going crazy because of I cannot catch it.
Reply | Threaded
Open this post in threaded view
|

Re: looking for a good owl pattern

Olivier Dameron
Hello,

On Tue, 24 Nov 2009 05:52:07 -0800 (PST), AlexJ <[hidden email]>
wrote:

> The most desirable way, from the first
> glance, would be to create a property hasKnowledgeArea and
> hasAlphabeticalView and then use it to connect ArticleIndividual to a
> particular Area. Right now I see only two way:
> - create an individual for each Knowledge Area and then connect
> article individual to KA individual
> - create a tree of defined abstract classes as subclasses of Article
> which will actually duplicate Knowledge Area tree so idividulas will
> be placed under right class by reasoner
>
> Both of them look like unsatisfied for me... I just wonder if someone
> was faced with the same problem and are there any patterns for that
> task?

Have you considered skos for example ?

cheers
olivier
_______________________________________________
protege-owl mailing list
[hidden email]
https://mailman.stanford.edu/mailman/listinfo/protege-owl

Instructions for unsubscribing: http://protege.stanford.edu/doc/faq.html#01a.03
Reply | Threaded
Open this post in threaded view
|

Re: looking for a good owl pattern

AlexJ
Thank you, Olivier,
but well:
    *  SKOS concepts are OWL individuals;
    * SKOS does not take a stance on whether it must also be possible to treat SKOS concepts as OWL classes;
    * The restrictions on OWL-DL prevent treating SKOS concepts as OWL classes;
    * There is an expectation that an ongoing OWL revision will alleviate the latter problem by offering some form of metamodelling.
http://www.w3.org/TR/2008/WD-skos-primer-20080221/

As far as I understand they tend to my second option ( to create an individual for each Knowledge Area class in terms of OWL). It is also mean then using SKOS which is not an OWL compliant, I will prevent myself from using reasoner, DL Querries/SPARQL etc, so what is the sense of using it? What benefits would I have if I do not have a possibility to use even Protege?
But the fact that SKOS exists indicated that I touched a big problem for OWL DL and it might be resolved only usinf OWL FUll, right? :), anyway, probably anybody has another solution adopting power of Querries and reasoner


<quite>
Have you considered skos for example ?

cheers
olivier


Reply | Threaded
Open this post in threaded view
|

Re: looking for a good owl pattern

Olivier Dameron
On Tue, 24 Nov 2009 06:46:11 -0800 (PST), AlexJ <[hidden email]>
wrote:

> As far as I understand they tend to my second option ( to create an
> individual for each Knowledge Area class in terms of OWL). It is also
> mean then using SKOS which is not an OWL compliant, I will prevent
> myself from using reasoner, DL Querries/SPARQL etc, so what is the
> sense of using it? What benefits would I have if I do not have a
> possibility to use even Protege?

ah, good point about skos and its limited relevance in your context.

I believe you could still use sparql, but you would have to give up the
owl-goodness

sorry for sending you on a dead end

olivier
_______________________________________________
protege-owl mailing list
[hidden email]
https://mailman.stanford.edu/mailman/listinfo/protege-owl

Instructions for unsubscribing: http://protege.stanford.edu/doc/faq.html#01a.03
Reply | Threaded
Open this post in threaded view
|

Re: looking for a good owl pattern

AlexJ
but now I know what SKOS is :) thank you for paying attention, it is a discussion and that is actually why we are here :)

Anyway, anybody HELP!!!


Olivier Dameron wrote
On Tue, 24 Nov 2009 06:46:11 -0800 (PST), AlexJ <alexjudin@gmail.com>
wrote

ah, good point about skos and its limited relevance in your context.

I believe you could still use sparql, but you would have to give up the
owl-goodness

sorry for sending you on a dead end

olivier
Reply | Threaded
Open this post in threaded view
|

Re: looking for a good owl pattern

samsontu
In reply to this post by AlexJ

Add, as superclasses of ArticleIndividual, "hasKnowledgeArea some
Boiling," "hasAlphabetView some A," etc.

Samson



AlexJ wrote:

> can anybody advise a good OWL pattern or at least point me at a right place:)
> for the following problem:
> there is a need to integrate some educational resources, lets call them
> Articles under different views like by Knowledge Areas, Alphabetical etc.
> So, I created the Classes as follows:
>
> Article
>   DeepProfessionalArticles
>   SurfaceEducationalArticles
>
> Knowledge Area
>   Thermodynamics
>      Boiling
>      etc
>   Heat and Mass Transfer
>      Heat Exchangers
>      etc
>
> Alphabetical View
>   A
>   B
>   C
>
> Now, I have to add individuals for Article, one by one, and then connect
> them to one or more Knowledge Areas. The most desirable way, from the first
> glance, would be to create a property hasKnowledgeArea and
> hasAlphabeticalView and then use it to connect ArticleIndividual to a
> particular Area. Right now I see only two way:
> - create an individual for each Knowledge Area and then connect article
> individual to KA individual
> - create a tree of defined abstract classes as subclasses of Article which
> will actually duplicate Knowledge Area tree so idividulas will be placed
> under right class by reasoner
>
> Both of them look like unsatisfied for me... I just wonder if someone was
> faced with the same problem and are there any patterns for that task? Or
> just point me to a right direction... The solution looks like to be on the
> surface and I am going crazy because of I cannot catch it.
>

_______________________________________________
protege-owl mailing list
[hidden email]
https://mailman.stanford.edu/mailman/listinfo/protege-owl

Instructions for unsubscribing: http://protege.stanford.edu/doc/faq.html#01a.03
Reply | Threaded
Open this post in threaded view
|

Re: looking for a good owl pattern

AlexJ
yep, this is my second solution (see below :) to duplicate all my Knowledge Tree by creating the same  structure but using Defined Abstract classes like you said... hasKnowledgeArea some Boiling then convert it to a defined class to make it a restriction. Well, but it is still unclear, when I am adding a new article individual, I have to say Article2 hasKnowledgeArea ???????  it should be an individual here... so anyway I have to create an invividual under each KnowledgeArea class , which seems like a triple work :)

So, it works, but the drawback it I have to do the same work twice, first create a Knowledge structure, then duplicate it as Article's subclasses and nevertheless I have to create individuals under KNArea classes. It does not look like an efficient solution :) Have I mistaken? Did I understand you correct?


Samson Tu wrote
Add, as superclasses of ArticleIndividual, "hasKnowledgeArea some
Boiling," "hasAlphabetView some A," etc.

Samson

> particular Area. Right now I see only two way:
> - create an individual for each Knowledge Area and then connect article
> individual to KA individual
> - create a tree of defined abstract classes as subclasses of Article which
> will actually duplicate Knowledge Area tree so idividulas will be placed
> under right class by reasoner
Reply | Threaded
Open this post in threaded view
|

Re: looking for a good owl pattern

samsontu
Sorry, I mis-spoke. I meant to say "Add, as *types* of
ArticleIndividual, "hasKnowledgeArea some Boiling," "hasAlphabetView
some A," etc.

I not NOT suggesting that you create defined (named) subclasses
"hasKnowledgeArea some Boiling" etc.

You DON'T say "Article2 hasKnowledgeArea some-instance." Instead, you
say Article2 is an instance of the anonymous class "hasKnowledgeArea
some Boiling," "hasAlphabetView some A," etc.

Samson


AlexJ wrote:

> yep, this is my second solution (see below :) to duplicate all my Knowledge
> Tree by creating the same  structure but using Defined Abstract classes like
> you said... hasKnowledgeArea some Boiling then convert it to a defined class
> to make it a restriction. Well, but it is still unclear, when I am adding a
> new article individual, I have to say Article2 hasKnowledgeArea ???????  it
> should be an individual here... so anyway I have to create an invividual
> under each KnowledgeArea class , which seems like a triple work :)
>
> So, it works, but the drawback it I have to do the same work twice, first
> create a Knowledge structure, then duplicate it as Article's subclasses and
> nevertheless I have to create individuals under KNArea classes. It does not
> look like an efficient solution :) Have I mistaken? Did I understand you
> correct?
>
>
>
> Samson Tu wrote:
>>
>> Add, as superclasses of ArticleIndividual, "hasKnowledgeArea some
>> Boiling," "hasAlphabetView some A," etc.
>>
>> Samson
>>
>>> particular Area. Right now I see only two way:
>>> - create an individual for each Knowledge Area and then connect article
>>> individual to KA individual
>>> - create a tree of defined abstract classes as subclasses of Article
>>> which
>>> will actually duplicate Knowledge Area tree so idividulas will be placed
>>> under right class by reasoner
>>
>

_______________________________________________
protege-owl mailing list
[hidden email]
https://mailman.stanford.edu/mailman/listinfo/protege-owl

Instructions for unsubscribing: http://protege.stanford.edu/doc/faq.html#01a.03
Reply | Threaded
Open this post in threaded view
|

Re: looking for a good owl pattern

Thomas Russ
In reply to this post by AlexJ

On Nov 24, 2009, at 5:52 AM, AlexJ wrote:

>
> can anybody advise a good OWL pattern or at least point me at a  
> right place:)
> for the following problem:
> there is a need to integrate some educational resources, lets call  
> them
> Articles under different views like by Knowledge Areas, Alphabetical  
> etc.
> So, I created the Classes as follows:
>
> Article
>  DeepProfessionalArticles
>  SurfaceEducationalArticles
>
> Knowledge Area
>  Thermodynamics
>     Boiling
>     etc
>  Heat and Mass Transfer
>     Heat Exchangers
>     etc
>
> Alphabetical View
>  A
>  B
>  C
>
> Now, I have to add individuals for Article, one by one, and then  
> connect
> them to one or more Knowledge Areas. The most desirable way, from  
> the first
> glance, would be to create a property hasKnowledgeArea and
> hasAlphabeticalView and then use it to connect ArticleIndividual to a
> particular Area. Right now I see only two way:
> - create an individual for each Knowledge Area and then connect  
> article
> individual to KA individual
> - create a tree of defined abstract classes as subclasses of Article  
> which
> will actually duplicate Knowledge Area tree so idividulas will be  
> placed
> under right class by reasoner

Well, if you don't mind having an OWL-Full ontology, you can also use  
the Classes as values of the properties on the individuals.  So you  
could then create an article, and link it directly to the class for  
its Knowledge Area.

As for the alphabetical view, I would prefer to compute that outside  
of the ontology (or possibly by using SWRL rules) rather than have to  
enter that information directly by hand.

_______________________________________________
protege-owl mailing list
[hidden email]
https://mailman.stanford.edu/mailman/listinfo/protege-owl

Instructions for unsubscribing: http://protege.stanford.edu/doc/faq.html#01a.03
Reply | Threaded
Open this post in threaded view
|

Re: looking for a good owl pattern

AlexJ
sorry for that gap in discussion :) it was a very long turkey coma )))
well, it is a big lure to use OWL Full and connect individuals directly to classes... but as far as I understand I would not be able to use reasoners ... forgive me for some fool questions but, starting from the very beginning, if I used OWL DL and reasoner, would I have any other benefits except having my OWL file consistent? And would I still be able to use Protege for maintaining my OWL?
Are there any other logical/match tools which might get some additional knowledge out of my OWL?


Thomas Russ wrote
On Nov 24, 2009, at 5:52 AM, AlexJ wrote:

>
> can anybody advise a good OWL pattern or at least point me at a  
> right place:)
> for the following problem:
> there is a need to integrate some educational resources, lets call  
> them
> Articles under different views like by Knowledge Areas, Alphabetical  
> etc.
> So, I created the Classes as follows:
>
> Article
>  DeepProfessionalArticles
>  SurfaceEducationalArticles
>
> Knowledge Area
>  Thermodynamics
>     Boiling
>     etc
>  Heat and Mass Transfer
>     Heat Exchangers
>     etc
>
> Alphabetical View
>  A
>  B
>  C
>
> Now, I have to add individuals for Article, one by one, and then  
> connect
> them to one or more Knowledge Areas. The most desirable way, from  
> the first
> glance, would be to create a property hasKnowledgeArea and
> hasAlphabeticalView and then use it to connect ArticleIndividual to a
> particular Area. Right now I see only two way:
> - create an individual for each Knowledge Area and then connect  
> article
> individual to KA individual
> - create a tree of defined abstract classes as subclasses of Article  
> which
> will actually duplicate Knowledge Area tree so idividulas will be  
> placed
> under right class by reasoner

Well, if you don't mind having an OWL-Full ontology, you can also use  
the Classes as values of the properties on the individuals.  So you  
could then create an article, and link it directly to the class for  
its Knowledge Area.

As for the alphabetical view, I would prefer to compute that outside  
of the ontology (or possibly by using SWRL rules) rather than have to  
enter that information directly by hand.

_______________________________________________
protege-owl mailing list
protege-owl@lists.stanford.edu
https://mailman.stanford.edu/mailman/listinfo/protege-owl

Instructions for unsubscribing: http://protege.stanford.edu/doc/faq.html#01a.03
Reply | Threaded
Open this post in threaded view
|

Re: looking for a good owl pattern

Thomas Schneider-5

On 4 Jan 2010, at 12:38, AlexJ wrote:

>
> sorry for that gap in discussion :) it was a very long turkey coma )))
> well, it is a big lure to use OWL Full and connect individuals  
> directly to
> classes... but as far as I understand I would not be able to use  
> reasoners
> ... forgive me for some fool questions but, starting from the very
> beginning, if I used OWL DL and reasoner, would I have any other  
> benefits
> except having my OWL file consistent?
Yes, you'd get all sorts of other consequences out of it:

* If you say that every duck is a bird and every bird is an animal,  
the reasoner would conclude that every duck is an animal. OK, this  
trivial entailment is already inferred without the use of a reasoner,  
but how about the next one:

* If you say that every duck has a wing and every wing is a body part,  
then the reasoner will conclude that every duck has a body part -- if  
you ask politely.

* If you say that every duck has a wing, that every wing contains  
tissue, and that the concatenation of "hasPart" and "contains" is a  
subproperty of "contains", then the reasoner will conclude that every  
duck contains tissue.

These entailments are still relatively simple. In general, entailments  
and finding their reasons can become so complex that separate  
justification services are needed (more pointers on request).

> And would I still be able to use
> Protege for maintaining my OWL?

Sure.

> Are there any other logical/match tools which might get some  
> additional
> knowledge out of my OWL?

There are. It depends on what kind of inferred knowledge you're after.

Cheers

Thomas

>
>
>
> Thomas Russ wrote:
>>
>>
>> On Nov 24, 2009, at 5:52 AM, AlexJ wrote:
>>
>>>
>>> can anybody advise a good OWL pattern or at least point me at a
>>> right place:)
>>> for the following problem:
>>> there is a need to integrate some educational resources, lets call
>>> them
>>> Articles under different views like by Knowledge Areas, Alphabetical
>>> etc.
>>> So, I created the Classes as follows:
>>>
>>> Article
>>> DeepProfessionalArticles
>>> SurfaceEducationalArticles
>>>
>>> Knowledge Area
>>> Thermodynamics
>>>    Boiling
>>>    etc
>>> Heat and Mass Transfer
>>>    Heat Exchangers
>>>    etc
>>>
>>> Alphabetical View
>>> A
>>> B
>>> C
>>>
>>> Now, I have to add individuals for Article, one by one, and then
>>> connect
>>> them to one or more Knowledge Areas. The most desirable way, from
>>> the first
>>> glance, would be to create a property hasKnowledgeArea and
>>> hasAlphabeticalView and then use it to connect ArticleIndividual  
>>> to a
>>> particular Area. Right now I see only two way:
>>> - create an individual for each Knowledge Area and then connect
>>> article
>>> individual to KA individual
>>> - create a tree of defined abstract classes as subclasses of Article
>>> which
>>> will actually duplicate Knowledge Area tree so idividulas will be
>>> placed
>>> under right class by reasoner
>>
>> Well, if you don't mind having an OWL-Full ontology, you can also use
>> the Classes as values of the properties on the individuals.  So you
>> could then create an article, and link it directly to the class for
>> its Knowledge Area.
>>
>> As for the alphabetical view, I would prefer to compute that outside
>> of the ontology (or possibly by using SWRL rules) rather than have to
>> enter that information directly by hand.
>>
>> _______________________________________________
>> protege-owl mailing list
>> [hidden email]
>> https://mailman.stanford.edu/mailman/listinfo/protege-owl
>>
>> Instructions for unsubscribing:
>> http://protege.stanford.edu/doc/faq.html#01a.03
>>
>>
>
> --
> View this message in context: http://n4.nabble.com/looking-for-a-good-owl-pattern-tp786837p998207.html
> Sent from the Protege OWL mailing list archive at Nabble.com.
> _______________________________________________
> protege-owl mailing list
> [hidden email]
> https://mailman.stanford.edu/mailman/listinfo/protege-owl
>
> Instructions for unsubscribing: http://protege.stanford.edu/doc/faq.html#01a.03
+----------------------------------------------------------------------+
|  Dr Thomas Schneider                    schneider (at) cs.man.ac.uk  |
|  School of Computer Science       http://www.cs.man.ac.uk/~schneidt  |
|  Kilburn Building, Room 2.114                 phone +44 161 2756136  |
|  University of Manchester                                            |
|  Oxford Road                                             _///_       |
|  Manchester M13 9PL                                      (o~o)       |
+-----------------------------------------------------oOOO--(_)--OOOo--+

Skagway (n.)
   Sudden outbreak of cones on a motorway.

                   Douglas Adams, John Lloyd: The Deeper Meaning of Liff





_______________________________________________
protege-owl mailing list
[hidden email]
https://mailman.stanford.edu/mailman/listinfo/protege-owl

Instructions for unsubscribing: http://protege.stanford.edu/doc/faq.html#01a.03

PGP.sig (210 bytes) Download Attachment
Reply | Threaded
Open this post in threaded view
|

Re: looking for a good owl pattern

AlexJ
well, what is always confusing me is how to link all of that to practical use... It looks like very atomizing and to be vety hard to extract new useful knowledge.
For instance, sticking to my original example (a number of articles linked to some Knowledge Areas), all I can infer from that ontology is the fact of belonging of an article to a branch of knowledge and that is all I can. I probably can also add some more properties like hasReference (Article hasReference Article) or hasAuthor (Article hasAuthor some Author).
What advantages may I have using DL here?  Linking to your Ducks and Wings, shall I add more detail like to split the boiling area to water, bubbles etc, wouldn't it be too much?
That is actually an answer on what kind of knowledge I am looking for... any not obvious knowledge which can not be easy inferred by user himself.

And thanks for your answers :)



Thomas Schneider-5 wrote
On 4 Jan 2010, at 12:38, AlexJ wrote:

>
> sorry for that gap in discussion :) it was a very long turkey coma )))
> well, it is a big lure to use OWL Full and connect individuals  
> directly to
> classes... but as far as I understand I would not be able to use  
> reasoners
> ... forgive me for some fool questions but, starting from the very
> beginning, if I used OWL DL and reasoner, would I have any other  
> benefits
> except having my OWL file consistent?

Yes, you'd get all sorts of other consequences out of it:

* If you say that every duck is a bird and every bird is an animal,  
the reasoner would conclude that every duck is an animal. OK, this  
trivial entailment is already inferred without the use of a reasoner,  
but how about the next one:

* If you say that every duck has a wing and every wing is a body part,  
then the reasoner will conclude that every duck has a body part -- if  
you ask politely.

* If you say that every duck has a wing, that every wing contains  
tissue, and that the concatenation of "hasPart" and "contains" is a  
subproperty of "contains", then the reasoner will conclude that every  
duck contains tissue.

These entailments are still relatively simple. In general, entailments  
and finding their reasons can become so complex that separate  
justification services are needed (more pointers on request).

> And would I still be able to use
> Protege for maintaining my OWL?

Sure.

> Are there any other logical/match tools which might get some  
> additional
> knowledge out of my OWL?

There are. It depends on what kind of inferred knowledge you're after.

Cheers

Thomas
Reply | Threaded
Open this post in threaded view
|

Re: looking for a good owl pattern

Thomas Schneider-5

On 4 Jan 2010, at 16:23, AlexJ wrote:

>
> well, what is always confusing me is how to link all of that to  
> practical
> use... It looks like very atomizing and to be vety hard to extract new
> useful knowledge.
> For instance, sticking to my original example (a number of articles  
> linked
> to some Knowledge Areas), all I can infer from that ontology is the  
> fact of
> belonging of an article to a branch of knowledge and that is all I  
> can. I
> probably can also add some more properties like hasReference (Article
> hasReference Article) or hasAuthor (Article hasAuthor some Author).
> What advantages may I have using DL here?  Linking to your Ducks and  
> Wings,
> shall I add more detail like to split the boiling area to water,  
> bubbles
> etc, wouldn't it be too much?
Well, your original example consists of a class hierarchy together  
with class/property assertions for individuals. Reasoning over such a  
structure is quite easy (almost boringly so) and doesn't need full-
fledged OWL reasoners. In fact, this structure only uses some of OWL's  
expressivity. As soon as you bring in more complex axioms (and my duck-
bird-etc example was only the beginning), ... reasoning becomes more  
complex. Whether you need more complex axioms depends on what you want  
to express. If a class hierarchy suffices, that's fine.

Cheers

Thomas

> That is actually an answer on what kind of knowledge I am looking  
> for... any
> not obvious knowledge which can not be easy inferred by user himself.
>
> And thanks for your answers :)


>
>
>
>
> Thomas Schneider-5 wrote:
>>
>>
>> On 4 Jan 2010, at 12:38, AlexJ wrote:
>>
>>>
>>> sorry for that gap in discussion :) it was a very long turkey  
>>> coma )))
>>> well, it is a big lure to use OWL Full and connect individuals
>>> directly to
>>> classes... but as far as I understand I would not be able to use
>>> reasoners
>>> ... forgive me for some fool questions but, starting from the very
>>> beginning, if I used OWL DL and reasoner, would I have any other
>>> benefits
>>> except having my OWL file consistent?
>>
>> Yes, you'd get all sorts of other consequences out of it:
>>
>> * If you say that every duck is a bird and every bird is an animal,
>> the reasoner would conclude that every duck is an animal. OK, this
>> trivial entailment is already inferred without the use of a reasoner,
>> but how about the next one:
>>
>> * If you say that every duck has a wing and every wing is a body  
>> part,
>> then the reasoner will conclude that every duck has a body part -- if
>> you ask politely.
>>
>> * If you say that every duck has a wing, that every wing contains
>> tissue, and that the concatenation of "hasPart" and "contains" is a
>> subproperty of "contains", then the reasoner will conclude that every
>> duck contains tissue.
>>
>> These entailments are still relatively simple. In general,  
>> entailments
>> and finding their reasons can become so complex that separate
>> justification services are needed (more pointers on request).
>>
>>> And would I still be able to use
>>> Protege for maintaining my OWL?
>>
>> Sure.
>>
>>> Are there any other logical/match tools which might get some
>>> additional
>>> knowledge out of my OWL?
>>
>> There are. It depends on what kind of inferred knowledge you're  
>> after.
>>
>> Cheers
>>
>> Thomas
>>
>>
>
> --
> View this message in context: http://n4.nabble.com/looking-for-a-good-owl-pattern-tp786837p998362.html
> Sent from the Protege OWL mailing list archive at Nabble.com.
> _______________________________________________
> protege-owl mailing list
> [hidden email]
> https://mailman.stanford.edu/mailman/listinfo/protege-owl
>
> Instructions for unsubscribing: http://protege.stanford.edu/doc/faq.html#01a.03
+----------------------------------------------------------------------+
|  Dr Thomas Schneider                    schneider (at) cs.man.ac.uk  |
|  School of Computer Science       http://www.cs.man.ac.uk/~schneidt  |
|  Kilburn Building, Room 2.114                 phone +44 161 2756136  |
|  University of Manchester                                            |
|  Oxford Road                                             _///_       |
|  Manchester M13 9PL                                      (o~o)       |
+-----------------------------------------------------oOOO--(_)--OOOo--+

Treewofe (n.)
   A very thick and heavy drift of snow balanced precariously on the
   edge of a door porch awaiting for what it judges to be the correct
   moment to fall.
   From the ancient Greek legend 'The Treewofe of Damocles'.

                   Douglas Adams, John Lloyd: The Deeper Meaning of Liff





_______________________________________________
protege-owl mailing list
[hidden email]
https://mailman.stanford.edu/mailman/listinfo/protege-owl

Instructions for unsubscribing: http://protege.stanford.edu/doc/faq.html#01a.03

PGP.sig (210 bytes) Download Attachment