KQL

[Alex Shkotin]

Colleagues,

The topic of Knowledged Query Language may be interesting exercise to follow SQL's [1] principles, components and sublanguages like this:

Knowledge Definition Language (where formal ontology may be seen)

Knowledge Manipulation Language (with "select" sublanguage)

Knowledge Constraints Language 

and even

Knowledge Procedural Language  

Knowledge Control Language

...

What nice terms!:-)

And, by the way, instead of Data Model we get Knowledge Model! (where KG may be seen)

What do you think?

Alex

[1] https://en.wikipedia.org/wiki/SQL

[John F. Sowa]

Alex,

Everybody on planet Earth knows and uses an excellent Knowledge Query Language every day.  It's our native language or some other NL that we choose to use or are required to use for some purpose.

But KQL is a bad acronym, because it puts too much emphasis on the Q.  It's better to emphasize D for dialogue than Q for question.  For an overview of the issues, see the slides (and references in them) about Cognitive Memory:  http://jfsowa.com/talks/cogmem.pdf

The cogmem.pdf slides discuss projects that our old VivoMind company implemented over a dozen years ago, and every one of those projects was specified by customers who paid for the implementations.  They weren't toy examples.  The bad news is that every one of them required a great deal of work by the VivoMind company to implement them.  There wasn't a single universal system that could be tailored by the customers themselves.

But there is newer technology that is customizable by anybody, not just the computer scientists.  I discussed that issue in a talk I presented at the Knowledge Graph Conference in 2020.  (By the way, it was awarded the Best Presentation prize.)  And I revised and extended it for the European Semantic Web Conference in June 2020.  I later added more slides from other presentations in http://jfsowa.com/talks/eswc.pdf

And by the way, I cc'd the Peirce email list because both the cogmem.pdf slides and the eswc.pdf slides build on the logic and semiotic of C.  S. Peirce.  I discuss that in some detail in the eswc.pdf slides.  Peirce called his existential graphs "the logic of the future", and he was right.

John

[Alex Shkotin]

John,

D is very important as I have written about "KnowledgeD Query Language" just to mimic "StructureD Query Language" whenever they called it this way:-)

The topic is about KQL components as we move from Data to Knowledge processing on planet Earth.

I hope we agree that KQL is a formal language. And if we look at formal English languages (some years ago there were 50+), you know my favorite  is ACE [1].

And you inspire me to the questions to Prof. Norbert Fuchs:

Dear Professor,

Where is in ACE:

-Knowledge Definitions Language?

- Knowledge Manipulation Language?

and so on, see [2].

Alex

[1] http://attempto.ifi.uzh.ch/site/

[2] https://groups.google.com/g/ontolog-forum/c/tT5VcfVqX-w/m/uJK4nbdpAgAJ

пн, 21 июн. 2021 г. в 05:52, John F. Sowa <so...@bestweb.net>:

--
All contributions to this forum are covered by an open-source license.
For information about the wiki, the license, and how to subscribe or
unsubscribe to the forum, see http://ontologforum.org/info/
---
You received this message because you are subscribed to the Google Groups "ontolog-forum" group.
To unsubscribe from this group and stop receiving emails from it, send an email to ontolog-foru...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/ontolog-forum/ee5285dd56f65060547fcf91e35c9da0.squirrel%40webmail2.bestweb.net.

[Alex Shkotin]

John, for you and other peirceanians,

It may be interesting today "16.30-17.00 Alex Belikov (MSU), Peirce's Triadic Logic and Its (Overlooked) Connexive Expansion"  Time zone is Moscow, RF.

see here https://llfp.hse.ru/en/formalphilosophy/2021/

Alex

пн, 21 июн. 2021 г. в 05:52, John F. Sowa <so...@bestweb.net>:

Alex,

--

[John F. Sowa]

AS> I hope we agree that KQL is a formal language. And if we look at formal English languages (some years ago there were 50+), you know my favorite  is ACE [1].

I don't agree.  Much more R&D has been done in the past dozen years in translating natural languages to formal notations.  They are not perfect, but they are good enough that the computer can provide an echo:  A translation from the estimated formal notation back to the same natural language from which it was derived.

Then the human can either say that the echo is correct or make some corrections.  That is much more user-friendly than just saying that the input is  not recognized.

Re ACE:  It was good for its day, but the technology has improved quite a bit since then.

John

[John Bottoms]

Agreed,

Further, the DARPA Lorelei project is in development proposes to translate from English to ~7000 languages, and back to English. It is corpus driven so there may not be sufficient corpora to support that total.

https://www.darpa.mil/program/low-resource-languages-for-emergent-incidents

-JohnB

--

All contributions to this forum are covered by an open-source license.
For information about the wiki, the license, and how to subscribe or
unsubscribe to the forum, see http://ontologforum.org/info/
---
You received this message because you are subscribed to the Google Groups "ontolog-forum" group.
To unsubscribe from this group and stop receiving emails from it, send an email to ontolog-foru...@googlegroups.com.

To view this discussion on the web visit https://groups.google.com/d/msgid/ontolog-forum/cd951b860cbefcea4d2a7a47126fc617.squirrel%40webmail2.bestweb.net.

[Alex Shkotin]

JFS,

There are some misunderstandings here:

-idea of KQL as in [1] is to be a formal language like SQL. If your K_Dialogue_L is an informal one this is very interesting but is up to you.

-Formal languages including ACE and KQL can be translated to other formal languages but this is not mandatory - just usual practice. For example we have a tableau algorithm for OWL2 and we translate ACE to OWL2 to use this algorithm. And for formal languages translation text1.fl1 to text2.fl2 and back to FL1 gives the same text text1.fl1 as far as I know.

Alex

[1] https://groups.google.com/g/ontolog-forum/c/tT5VcfVqX-w/m/uJK4nbdpAgAJ

вт, 22 июн. 2021 г. в 06:45, John F. Sowa <so...@bestweb.net>:

--

All contributions to this forum are covered by an open-source license.
For information about the wiki, the license, and how to subscribe or
unsubscribe to the forum, see http://ontologforum.org/info/
---
You received this message because you are subscribed to the Google Groups "ontolog-forum" group.
To unsubscribe from this group and stop receiving emails from it, send an email to ontolog-foru...@googlegroups.com.

To view this discussion on the web visit https://groups.google.com/d/msgid/ontolog-forum/cd951b860cbefcea4d2a7a47126fc617.squirrel%40webmail2.bestweb.net.

[Alex Shkotin]

JohnB,

This sounds great "The technologies resulting from LORELEI research will be capable of supporting situational awareness based on low-resource foreign language sources within an extremely short time frame – starting as soon as 24 hours after a new language requirement emerges." for 2014. Is there any result for 2021?

Alex

вт, 22 июн. 2021 г. в 06:58, John Bottoms <jo...@firststarsystems.com>:

To view this discussion on the web visit https://groups.google.com/d/msgid/ontolog-forum/f1b9f570-6e63-b70d-aceb-89701abd3360%40firststarsystems.com.

[alex.shkotin]

And let me add Knowledge Management Systems with respect to DBMS.

Do you know one except Protege for OWL2, like Oracle DBMS or PostgreSQL for SQL?

Well, maybe it should sounds like KG MS:-) or even Ontology MS!

воскресенье, 20 июня 2021 г. в 10:00:45 UTC+3, alex.shkotin:

[William Frank]

I want to echo John's point in his first message of this thread: "It's better to emphasize D for dialogue than Q for question" 

'Query' is something you can do to a database', there are the four common CRUD speech acts one can perform.

Knowledge, on the other hand,  results from dialogs.  It is not inert like a piece of data.  Knowledge is embodied in the justified beliefs of a community that works together on extending and refining and revising their knowledge of some domain,  The activities one engages in to pursue knowledge are things like explaining, teaching, questioning, imagining examples and counterexamples, disagreeing, examining a body of assertions, deducing, finding contradictions. Knowledge bases, like the corpus of books and papers and meetings on bacgterfial diseases, or the Case Law of the New York superior courts, or the body of knowledge of effective procedures, regulations, and observations that a company evolves collaboratively.  

Knowledge is not a data sheep in wolf's clothing - it's a completely different animal.   "Wait a minute, I'll ask the knowledge base" ??? you couldn't put assertions in a real knowledge base the way you can put data in a database, Knowlege  is something  some people have, that results from their making and modifying informed decisions when someone dips into the Brownian motion of the community's interactions, 

So, taking a lot of database terms, like query language, data definition language, and replacing the word 'data' with the word 'knowledge'  is unlikely in my opinion to bear much fruit.  And an ontology graph with a bunch of instance facts hung off it is, in my opinion, a very useful thing and an advance of our practice, but is a database. its metadata, a means of making inferences. and the claim and evidence that the ontology graph represents the concept map of a domain, and the some kind of verification of the asserted facts beyond some software putting things in a database.  But 'knowledge base' is a bit grand for that small increment of practice.  

On the other hand, when Microsoft 'invented' a proprietary LDAP server,  they gave it a different name,.    If Bill could and did get away with that,  we can *call* that new thing we do how to build a 'knowledge base".  To keep this from going to our heads, though, and thinking we've 'done'  knowledge instead of just having started, it might be well to recall Abraham Lincoln's anti-humpty dumpty stance:  "If you call a tail a leg, how many legs does a dog have?  Still only four; calling a tail a leg don't make it one."     

To view this discussion on the web visit https://groups.google.com/d/msgid/ontolog-forum/CAFxxROQsUfqgZKK3HjaC%2B4ZDOgePZc5d39bvP9yoUr10%3Dw%3DJPA%40mail.gmail.com.

--

Wm

We understand what other people say through empathy—imagining ourselves to be in the situation they were in, including imaging wanting to say what they wanted to say.  

– Zellig Harris    

[Alex Shkotin]

Wm,

I am not about Q itself, I am about 

"Knowledge Definition Language (where formal ontology may be seen)

Knowledge Manipulation Language (with "select" sublanguage)

Knowledge Constraints Language 

and even

Knowledge Procedural Language  

Knowledge Control Language

...

Knowledge Model (where KG may be seen)

...

Knowledge Management System

"

Technologies to work with Knowledge should be more subtle then with Data one?

Please, look at this like that: we have Data Manipulation Language. Do we need  Knowledge Manipulation one?

etc.

Alex

ср, 23 июн. 2021 г. в 08:44, William Frank <william...@gmail.com>:

To view this discussion on the web visit https://groups.google.com/d/msgid/ontolog-forum/CALuUwtC-jHjsNbnxxszsOYrQ_h84X%3DZsd_T-1g2M%2BwoaJF5gUg%40mail.gmail.com.

[Alex Shkotin]

And let me cite "Gartner defines knowledge workers are those who are involved in knowledge-intensive occupations, such as writers, accountants and engineers." from just arrived

https://www.techrepublic.com/article/gartner-51-of-global-knowledge-workers-will-be-remote-by-the-end-of-2021/

ср, 23 июн. 2021 г. в 08:44, William Frank <william...@gmail.com>:

I want to echo John's point in his first message of this thread: "It's better to emphasize D for dialogue than Q for question" 

To view this discussion on the web visit https://groups.google.com/d/msgid/ontolog-forum/CALuUwtC-jHjsNbnxxszsOYrQ_h84X%3DZsd_T-1g2M%2BwoaJF5gUg%40mail.gmail.com.

[William Frank]

Hi, Alex, 

I am not objecting to what you are doing, only to what our emerging profession seems to support with its own hype the end of freedom of thought, belief, and action. 

The underlying reason for my view is that 

• More and more people accept that if S is an assertion they find in the computer, it is true.  Instead of "We have been unable to find your account records in our system.", it's now: "You don't have an account with us."
• If we encourage people to start to call what they find in the computer 'knowledge",  the game is over.  

So, I am sad to see that this is what is happening - that people will be calling what "the oracle computer" tells them knowledge. 

I would just be happier if there were some other more accurate word used for what you are working on.  

But calling it knowledge is dangerous: What's in a word?  A whole world of persuasion and attitude? 

 - 

So, when I say I do not think that knowledge can be defined, controlled, manipulated, constrained. etc., thus languages that purport to do so are doing something else, I am arguing for a change of word choice to describe this kind of work. ( Though, I have no expectation that such a change will come about.)  

To view this discussion on the web visit https://groups.google.com/d/msgid/ontolog-forum/CAFxxROTu5FPEzNya7%3D2LpRA3MS%3D_8bOfvTaV04ot9G6D0JgiJQ%40mail.gmail.com.

[Alex Shkotin]

Wm,

Knowledge processing is a busy term [1] just google it. My point is only to bring formal axiomatic systems and their model techniques to the area of formal ontology and KG.

At least from the KIF time we know that we can exchange knowledge:-)

Alex

[1] https://www.sciencedirect.com/topics/computer-science/knowledge-processing

ср, 23 июн. 2021 г. в 16:12, William Frank <william...@gmail.com>:

To view this discussion on the web visit https://groups.google.com/d/msgid/ontolog-forum/CALuUwtAJwed1vx%3DfTFtSTnCRFK4Boj%3DYg8ZygjKxwnmhAgGRdA%40mail.gmail.com.

[William Frank]

On Wed, Jun 23, 2021 at 12:31 PM Alex Shkotin <alex.s...@gmail.com> wrote:

Wm,

Knowledge processing is a busy term [1] just google it.

 

Yes, it is a busy term.    That is my very point, that this use of the term is perverting the world view of many of those who hear it or use it, into one in which knowledge is something that could be stored or 'processed' as something that looks a lot like data.    The broad use of a term does not mean it is a use that will improve humanity. OTOH, I do not expect that use to go away, or that people who are aware of its dangers must avoid using it that way, thus removing themselves from the language of the community in which they are working.  

My point is only to bring formal axiomatic systems and their model techniques to the area of formal ontology and KG.

And that is a very important thing to do, something I have supported and called for two decades, .

 

At least from the KIF time we know that we can exchange knowledge:-)

The KIF was a very good thing too.  (I think KIF is an ancestor of common logic?)  

Only, these enable computers to share assertions, assertions that might be intentional lies, or just asserted without justification, for example.   If it just so happens that some assertion stored in some computer represents sound knowledge, then it can exchange that.  The Kims of North Korea,  could have a"knowledge base" full of things such as how people can achieve a certain kind of immortality by participating in the state and aligning themselves with their immortal leaders.

People can exchange anything using common logic from their ungrounded opinions (most popular) to the best evidenced and theoretically consistent assertions about a domain.   They can exchange these, and some of what they exchange might in context represent some knowledge.    But that has nothing to do with the fact that it is expressed in Common Logic or Italian.

Thanks for pursuing this a bit. 

Wm

To view this discussion on the web visit https://groups.google.com/d/msgid/ontolog-forum/CAFxxROT%2Bg_TBgUPKj-0m_HzD5vGGrQUz_gCo3xgn9tUWBqptUg%40mail.gmail.com.

[Alex Shkotin]

Wm,

Just to add a little bit to the knowledge of what they are doing with knowledge. This is from our conference yesterday: "Knowledge and Time.

It is important to formalize the crucial idea that agent's knowledge evolves over time due to some events.
Many logics describe the dynamics of knowledge change (DEL, PAL, ETL, KL(n), TEL).

Nevertheless, they do not allow us to formalize particular kind of scenarios of knowledge changes over time." [1]

Sounds like OMG:-)

Alex

[1] https://prezi.com/p/ackzp-ddoyi0/formal-philosophy-2021/

чт, 24 июн. 2021 г. в 00:08, William Frank <william...@gmail.com>:

To view this discussion on the web visit https://groups.google.com/d/msgid/ontolog-forum/CALuUwtDJcu%3DURg%3D4p6-T%3DfzvnyNP9gr%3DSm%3DyV4xewXppHeSHgA%40mail.gmail.com.