Re: [ontolog-forum] API4KB and diverse ontologies

[Ray Martin <marsaviator@xxxxxxxxx>]

Elisa,

 The RFP at http://www.omgwiki.org/API4KB/lib/exe/fetch.php?media=api4kb:rfp.pdf , section 6.1, para 2, says "These APIs are brittle, ..."

What is the meaning that was intended here? In what sense are the mentioned APIs brittle?

I am not necessarily expecting you to answer this, but I was pondering what are the expectations of the proposal for API4KB to overcome this brittleness?

I was also wondering if the current set of proposals have been able to address this brittleness?

thanx,

Ray

On Sat, Jun 29, 2013 at 3:05 PM, Elisa Kendall <ekendall@xxxxxxxxxxxx> wrote:

Hi Ray,

In API4KBs, our requirements include a broader range of interfaces (e.g., enterprise service bus), with targets that may include, but are not focused on CMS systems.  We have been reviewing what they did, and continue to be informed by it, but concluded that it didn't cover our use cases.   One possible outcome is that we may suggest extensions to their work to provide a compliant implementation, if they are open to that, when the time comes.

For example, one of the primary use cases for our work is to replace some of the underpinnings of the CTS2 2 set of standards from the OMG with more general purpose interfaces.  CTS 2 is publically available on the OMG site, although I'm not sure if recent updates have been posted yet.  The CTS 2 effort represents joint work between HL7 and the OMG, and provides terminology services for healthcare, largely developed at the Mayo Clinic but with other participants as well.  The architecture has been in place for many years, although the standard has only been available in the last couple of years from the OMG.

Architectures that are motivating our work include not only CTS 2 and its predecessors, but the Reaction RuleML effort, which has been well tested with commercial rule systems and other applications such as analytics, but without any description logics reasoning in the loop.  We have also gone back to the FIPA work, and especially their ACL language approach, to facilitate looser coupling, as our goal is to create more of a framework than tightly coupled interfaces, with specific implementations to support a couple of key use cases and provide a reference for others.

I hope this provides a bit more clarity,

Elisa

On 6/29/2013 10:09 AM, Ray Martin wrote:

here are the listed components of Stanbol:

https://stanbol.apache.org/docs/trunk/components/

quick list:

• Enhancer
  • NLP processing
  • List of Enhancement Engines
  • Enhancement Structure
• Entityhub
• Contenthub
• Ontology Manager
• Rules
• Reasoners
• CMS Adapter
• FactStore

From 

http://www.omgwiki.org/API4KB/doku.php

here is their list:

• Parsing
• Reasoning
• Persistent Storage
• Queries
• Rules

Certainly some of the same wording. Takes a lot of study to discover the intent and meaning behind the words.

As one attempts to produce a system, one realizes that one will be thousands of years old if one produces all components by one's lonesome.  So, one attempts to use ideas and accomplishments of others - Stanbol, API4KB, etc, etc - only to discover that because humans cannot work together, one will be thousands of years old before knowing what, how, when, and where to utilize the multitude of efforts.  It is indeed very nice that folks opensource their efforts - but, WOW??? ultra-confusing...

On Sat, Jun 29, 2013 at 9:34 AM, John F Sowa <sowa@xxxxxxxxxxx> wrote:

On 6/29/2013 8:19 AM, Ray Martin wrote:
> I am trying to understand the similarities and differences between
> API4KB and Apache Stanbol. If I select the mechanisms of one project
> will the difficulties of using the other framework be insurmountable?

Good question.

Since the API4KB files are password protected, I can't say anything
about it beyond the slides.  But Apache has various projects that
were started by various groups and donated to the Apache Foundation
after they were fairly well developed.

 From Wikipedia:
> Apache Stanbol is an open source modular software stack and reusable
> set of components for semantic content management. Apache Stanbol
> components are meant to be accessed over RESTful interfaces to
> provide semantic services for content management. Thus, one
> application is to extend traditional content management systems
> with (internal or external) semantic services.

 From https://stanbol.apache.org/
> Functionalities are provided as RESTful services returning results
> as RDF (Resource Description Language) and JSON. Apache Stanbol
> also supports the use of JSON-LD.

The term 'semantic services' suggests logic, but the quotation above
indicates that they don't go beyond RDF and JSON in expressive power.
Since the API4KB slides show that they're interested in reasoning,
they should have more expressive power.

But I get *very uneasy* when anybody starts to talk about reasoning
and does not say anything about logic.  In the DAML proposal of 2000,
Tim B-L presented SWeLL as a highly expressive foundational logic
(superset of propositional logic, FOL, and HOL).  Hayes and Guha
designed LBase in 2003 as a logical foundation consistent with both
Tim's SWeLL and Common Logic.

That could have been (and still could be) the foundation for all
of these components.  I have no complaints about a multiplicity
of different APIs that are tailored for each kind of syntax.
But they should all be related to a common semantics.

John

_________________________________________________________________
Message Archives: http://ontolog.cim3.net/forum/ontolog-forum/
Config Subscr: http://ontolog.cim3.net/mailman/listinfo/ontolog-forum/
Unsubscribe: mailto:ontolog-forum-leave@xxxxxxxxxxxxxxxx
Shared Files: http://ontolog.cim3.net/file/
Community Wiki: http://ontolog.cim3.net/wiki/
To join: http://ontolog.cim3.net/cgi-bin/wiki.pl?WikiHomePage#nid1J

 
_________________________________________________________________
Message Archives: http://ontolog.cim3.net/forum/ontolog-forum/  
Config Subscr: http://ontolog.cim3.net/mailman/listinfo/ontolog-forum/  
Unsubscribe: mailto:ontolog-forum-leave@xxxxxxxxxxxxxxxx
Shared Files: http://ontolog.cim3.net/file/
Community Wiki: http://ontolog.cim3.net/wiki/ 
To join: http://ontolog.cim3.net/cgi-bin/wiki.pl?WikiHomePage#nid1J
 

_________________________________________________________________
Message Archives: http://ontolog.cim3.net/forum/ontolog-forum/
Config Subscr: http://ontolog.cim3.net/mailman/listinfo/ontolog-forum/
Unsubscribe: mailto:ontolog-forum-leave@xxxxxxxxxxxxxxxx
Shared Files: http://ontolog.cim3.net/file/
Community Wiki: http://ontolog.cim3.net/wiki/
To join: http://ontolog.cim3.net/cgi-bin/wiki.pl?WikiHomePage#nid1J
 

_________________________________________________________________
Message Archives: http://ontolog.cim3.net/forum/ontolog-forum/  
Config Subscr: http://ontolog.cim3.net/mailman/listinfo/ontolog-forum/  
Unsubscribe: mailto:ontolog-forum-leave@xxxxxxxxxxxxxxxx
Shared Files: http://ontolog.cim3.net/file/
Community Wiki: http://ontolog.cim3.net/wiki/ 
To join: http://ontolog.cim3.net/cgi-bin/wiki.pl?WikiHomePage#nid1J    (01)