Re: Applications

Sebastian

Thank you

Maybe writing/publishing  a paper with the ideas so far motivations and
whatever you have - could be useful to present the thinking in some
coherent way to those who may want to stigmergise with the project

I have half a dozen applications/frameworks in the works at any given time
which I dont have the time and resources to develop.

It depends where you want to take these ideas, whether they becomes
something its up to
you,  Just a paper with the thinking is fine as well,  If you want to
develop it into a project application you need more effort, commitment and
resources/people  etc,  Maybe present your stuff at a conference and see if
something sticks,

I cannot help with the RDF quads input sorry- if the work you are doing is
related to the mission of this CG
you may have to come up with some concrete proposal

In terms of KR, a lot can be demonstrated/presented on paper,  with minimal
proof of concept

What becomes of our ideas depends on how convinced we are, how much time we
have
to devote to them etc.

These days I value going for a walk and a swim on the beach more than
convincing anyone about anything
in particular

PDM




<https://www.avast.com/sig-email?utm_medium=email&utm_source=link&utm_campaign=sig-email&utm_content=webmail&utm_term=icon>
Virus-free.
www.avast.com
<https://www.avast.com/sig-email?utm_medium=email&utm_source=link&utm_campaign=sig-email&utm_content=webmail&utm_term=link>
<#DAB4FAD8-2DD7-40BB-A1B8-4E2AA1F9FDF2>

On Sun, Apr 28, 2019 at 6:57 AM Sebastian Samaruga <ssamarug@gmail.com>
wrote:

> Paola, I'll try to respond you inline:
>
> On Mon, Apr 22, 2019, 1:30 AM Paola Di Maio <paoladimaio10@gmail.com>
> wrote:
>
>> Sebastian
>> thanks for sharing
>> I understand better now
>>
>> The reason why we communicate research is often to clarity it to
>> ourselves first, as well as to get it appreciated/rubbished by others. :-)
>>
>> I see two possible points of interest (correct if wrong)
>> 1. the use of RDF  quads (this could demand a use case type of research
>> method)
>> *(how does this relate to the rest of the stuff?)*
>>
>
> I've initially tried to tackle the problem of ontology matching: identity
> resolution between different 'names' meaning the same entity in some
> context, using 'semiotic' (sign, concept, object) aggregation / relations
> for that.
>
> Then, my use of RDF quads developed into a form of (hierarchy layered)
> object graph serialization mechanism. Then, the 'layers' mechanism showed
> to allow some kind of hierarchical aggregation and allow to 'infer': data,
> schema and behavior 'layers' of an application by inspecting its data CRUD
> messages (plain CSPO statements).
>
>
>> 2. the development of a new construct/mechanism/representation
>> To me these two seem possibly distinct
>>
>
> In fact, this is what could happen when a functional / object oriented
> implementation exists. My plan is to abstract URI 'Connectors' implementing
> event driven sources / sink mechanisms (HTTP verbs wrapping a database, for
> example) and having a Functional Resource abstraction which homologue
> backends.
>
>
>> For either or both, it would be good to understand the novelty, the
>> benefits (what problem do es the novelty solve? as opposed to not using it,
>> or using other approaches)
>>
>
> The principal outcome of this proposed approach is to enable developers
> for having a suite of tools, kind of framework, for augmenting, extending
> or creating applications.
>
> This, having abstractions of much higher level of plain current RDF stores
> APIs providing a set of pluggable components, like the ones existing for
> other plataforms (Java EE, Spring, Drupal, etc.).
>
> then you have to demonstrate your claim with some experiment or example
>> you then need to validate the outcome of experiment of example
>>
>
> I'm currently far far away of having anything implemented in any platform.
> Just documents. I've just uploaded 'Index3.docx' but it's just an index /
> roadmap for previous scrapbook.
>
> I'd really appreciate if someone can help me with the 'Introduction to RDF
> Graphs, Triples, Quads' section.
>
>
>> If you could draft a few paragraphs to address each of the points
>> then you can put your work up for peer review, I like open peer reviews
>> Maybe ask some academic on this list to pre-review
>>
>> The you can submit to some publication of your choice
>>
>
> I'm sorry if what I currently have is just a set of thoughts in an
> scrapbook. Any kind of help will be greatly appreciated. As you said, I
> share in the hope of clarify stuff.
>
> Best Regards,
> Sebastián.
>
>
>
>>
>> P
>>
>>
>> On Sat, Apr 20, 2019 at 1:12 AM Sebastian Samaruga <ssamarug@gmail.com>
>> wrote:
>>
>>> Paola,
>>>
>>> I'll try to answer to your questions inline. But first, let me tell you
>>> that the links I've posted on my first mail have some updates (blog and
>>> documents):
>>>
>>> https://github.com/snxama/scrapbook/blob/master/Index2.docx?raw=true
>>>
>>> https://github.com/snxama/scrapbook/blob/master/Application.docx?raw=true
>>>
>>> http://snxama.blogspot.com
>>>
>>>
>>> On Thu, Apr 18, 2019, 11:30 PM Paola Di Maio <paoladimaio10@gmail.com>
>>> wrote:
>>>
>>>> Sebastian
>>>> thank a lot, I can see you speak geek  :-)
>>>>
>>>> Is it a new method that you are proposing?
>>>> I would find useful to understand if you could highlight
>>>> - how does it relate to mechanisms already in place (I suspect your
>>>> approach builds and extends on  existing methods?)
>>>>
>>>
>>> In facts, I'm planning to use RDF Quads in a non "canonical" way. Maybe
>>> I'll be using a RDF store for local persistence and DIDs (Distributed
>>> Identifiers) for decentralized persistence of a distributed network.
>>>
>>> https://w3c-ccg.github.io/did-spec/
>>>
>>> Given that one of my goals is to infer applications data / schema /
>>> behavior from the raw statements they could produce or consume for being
>>> able to match this ontologies (data / schema / behavior matchings) from
>>> different sources / applications and provide a mechanism for integration,
>>> I'd be using a set of "connector" URIs implementations (events sources /
>>> sinks) implementing whatever protocols / back ends I'd like to "plug" in
>>> the network.
>>>
>>> So, it builds and extends on existing methods as long as one could write
>>> an URI connector for that method: databases, services, SPARQL endpoints,
>>> etc.
>>>
>>> - the benefits (what does this approach do that existing approaches
>>>> don't?)
>>>>
>>>
>>> URI implementations (connectors) will provide mechanisms for synchronize
>>> back ends / data sources with the augmented view of the interactions
>>> performed with this framework and interacting directly with this framework
>>> through API protocols will enable to declaratively build applications and
>>> services on top and extending existing augmentations.
>>>
>>> - limitations
>>>>
>>>
>>> As long as this is, almost now, only and analysis and design draft,
>>> implementation concerns are of mayor importance.
>>>
>>> - examples of applications with some kind of evaluation of the benefits
>>>> and limitations
>>>>
>>>
>>> In one of the documents I've sent I describe a potential "social"
>>> Producs And Services Community Exchange Network. Maybe SoLiD (
>>> solid.mit.edu) or something similar could solve the "social" part. And
>>> goods exchange is to be orchestrated by "smart" matching of requesters /
>>> offerers
>>>
>>> -  is this something you would want to publish as a concept or suggest
>>>> for adoption by W3C or our working group in any way (if so say how you
>>>> envisage)
>>>>
>>>
>>> Yes, In the realm of programming and ontology design patterns.
>>>
>>>
>>>> When you have this info I wonder if this could make a useful submission
>>>> for our forthcoming special issue AIKR, and in general it is good to send
>>>> stuff for peer review
>>>> unless this is going to become a patent or something
>>>>
>>>
>>> Think it could be. As long as I can make a clear and understandable
>>> document of all this.
>>>
>>>
>>>> PDM
>>>>
>>>> On Fri, Apr 19, 2019 at 3:14 AM Sebastian Samaruga <ssamarug@gmail.com>
>>>> wrote:
>>>>
>>>>> Paola, thanks for the feedback.
>>>>>
>>>>> What my attempts are about where, in the beginning, to match different
>>>>> URIs or identifiers which refer to the same entity (in different databases
>>>>> / ontologies, for example) to perform some kind of "ontology matching".
>>>>>
>>>>> Then I've tried to develop a mechanism for using RDF Quads for
>>>>> encoding an object graph (and a layers class hierarchy) using Contexts to
>>>>> denote the class of an instance, Subjects to denote class instances and
>>>>> attributes (members) and values: Predicates / Objects.
>>>>>
>>>>> Quads are "reified" as Resource(s). Also, Resource is a functional
>>>>> wrapper reactive and event driven of an URI. And an URI could be
>>>>> implemented with whatever backend which could produce or consume events
>>>>> (databases, services, etc.). Resource layers hierarchy (Context) is to be
>>>>> implemented by an actor / role type object pattern.
>>>>>
>>>>> Then I've realized that some basic type inference could be performed
>>>>> with, for example, aggregating Subjects with the same predicates (Subject
>>>>> Kinds). Idem for Predicates, Objects and Contexts. I've also realized that
>>>>> plain "facts" statements could be aggregated in the previously mentioned
>>>>> class hierarchy to abstract further, from plain data, instance / class
>>>>> layers of what I call data / schema / behavior layers. Higher layers (i.e.:
>>>>> Behavior) "aggregate" lower layers.
>>>>>
>>>>> Layers shape is as follow:
>>>>>
>>>>> Resource : Functional URI wrapper.
>>>>>
>>>>> (Context : Resource, Occurrence : Resource, Attribute : Resource,
>>>>> Value : Resource);
>>>>>
>>>>> Each layer abstract:
>>>>>
>>>>> Statement (data instance):
>>>>>
>>>>> (Statement, Occurrence, Attribute, Value);
>>>>>
>>>>> someOne buys someProduct;
>>>>>
>>>>>
>>>>> Entity (data class):
>>>>> (Entity, Statement, Occurrence, Attribute);
>>>>> someBuyer, someProduct (Entity);
>>>>>
>>>>> Role (schema instance):
>>>>> (Role, Entity, Statement, Occurrence);
>>>>> Buyer, Product (Role);
>>>>>
>>>>> Class (schema class):
>>>>> (Class, Role, Entity, Statement);
>>>>> Person, Good (Class);
>>>>>
>>>>> Flow (behavior instance):
>>>>> (Flow, Class, Role, Entity);
>>>>> someBought (Flow);
>>>>>
>>>>> Behavior (behavior class):
>>>>>
>>>>> (Behavior, Flow, Class, Role);
>>>>> Buy (Behavior);
>>>>>
>>>>> This "aggregations" are part of what I call "Augmentation(s)":
>>>>> Aggregation, Alignment and Activation are ones of those, which are
>>>>> functional transforms described declaratively in an object graph metamodel.
>>>>> The act of applying an Augmentation implies one source Resource (context),
>>>>> one template Resource (transform) and a resulting (set of) Resource(s).
>>>>>
>>>>> One also could Augment Resource(s) in a functional manner, using
>>>>> reactive event driven APIs so, for example applying "Person" class to
>>>>> "Employee" role could shield a Resource set of people being working for
>>>>> someone. The ultimate goal is to be able to "plug" as much "backends"
>>>>> connectors as posible into distributed peers which exposes protocols / APIs
>>>>> for knowledge driven hypermedia applications.
>>>>>
>>>>> Best Regards,
>>>>> Sebastián.
>>>>>
>>>>>
>>>>> On Sun, Apr 14, 2019, 9:48 PM Paola Di Maio <paola.dimaio@gmail.com>
>>>>> wrote:
>>>>>
>>>>>> Sebastian
>>>>>>
>>>>>> thanks for the post  and for sharing your docs
>>>>>> I have only briefly glanced through them - working on deadlines all
>>>>>> the time -
>>>>>> and it looks interesting.
>>>>>> I think many of us struggle to be coherent and no, this forum primary
>>>>>> goal is not  psychoanalysis although many of us could do with some :-)
>>>>>>
>>>>>> Sharing thoughts can help us better formulate questions
>>>>>> \
>>>>>> I would be grateful if you could explain a bit more of  what you are
>>>>>> sharing with us, and what kind of input you d like to receive on the docs,
>>>>>> or explain how these can contribute to the mission
>>>>>>
>>>>>> thank you,
>>>>>>
>>>>>> PDM
>>>>>>
>>>>>>
>>>>>>
>>>>>> On Fri, Apr 12, 2019 at 2:43 AM Sebastian Samaruga <
>>>>>> ssamarug@gmail.com> wrote:
>>>>>>
>>>>>>> Sorry if I dare to post my spare thoughts again in this lists. I
>>>>>>> apologize but it is in the hope of sharing what comes into my writings
>>>>>>> trying to look for someone with more knowledge to "validate" or to "guide"
>>>>>>> my points of view.
>>>>>>>
>>>>>>> Going through my most recent attempts of having something concrete
>>>>>>> for sharing in plain English I realize one mistake I'm committing: I'm
>>>>>>> trying to describe combustion vehicles (Hypermedia Applications) saying
>>>>>>> that petroleum exists (Semantic Intelligence).
>>>>>>>
>>>>>>> As long as my post are going I've just got a stack of (incoherent)
>>>>>>> "analysis" documents as the result of my work. And I had only those until
>>>>>>> now because I was stuck because of the previously mentioned mistake (ah,
>>>>>>> and because of my Bipolar Disease maniac episodes...).
>>>>>>>
>>>>>>> So, I should try to describe applications instead and see how and
>>>>>>> where fuel should burn properly inside a motion vehicle to generate
>>>>>>> traction. Every semicolon I write is updated into my GitHub repository, so,
>>>>>>> sorry if you browse that "scrapbook" and you don't find anything even
>>>>>>> intelligible.
>>>>>>>
>>>>>>>
>>>>>>> https://github.com/snxama/scrapbook/blob/master/Application.docx?raw=true
>>>>>>>
>>>>>>> There is a brief technical description work in progress document.
>>>>>>> For now it just a list of statements about a potential backend. Session and
>>>>>>> Inteteraction layers are not specified.
>>>>>>>
>>>>>>> https://github.com/snxama/scrapbook/blob/master/Index2.docx?raw=true
>>>>>>>
>>>>>>> Best Regards,
>>>>>>> Sebastian Samaruga:
>>>>>>> http://snxama.blogspot.com
>>>>>>>
>>>>>>>
>>>>>
>>>>> On Sun, Apr 14, 2019, 9:48 PM Paola Di Maio <paola.dimaio@gmail.com>
>>>>> wrote:
>>>>>
>>>>>> Sebastian
>>>>>>
>>>>>> thanks for the post  and for sharing your docs
>>>>>> I have only briefly glanced through them - working on deadlines all
>>>>>> the time -
>>>>>> and it looks interesting.
>>>>>> I think many of us struggle to be coherent and no, this forum primary
>>>>>> goal is not  psychoanalysis although many of us could do with some :-)
>>>>>>
>>>>>> Sharing thoughts can help us better formulate questions
>>>>>> \
>>>>>> I would be grateful if you could explain a bit more of  what you are
>>>>>> sharing with us, and what kind of input you d like to receive on the docs,
>>>>>> or explain how these can contribute to the mission
>>>>>>
>>>>>> thank you,
>>>>>>
>>>>>> PDM
>>>>>>
>>>>>>
>>>>>>
>>>>>> On Fri, Apr 12, 2019 at 2:43 AM Sebastian Samaruga <
>>>>>> ssamarug@gmail.com> wrote:
>>>>>>
>>>>>>> Sorry if I dare to post my spare thoughts again in this lists. I
>>>>>>> apologize but it is in the hope of sharing what comes into my writings
>>>>>>> trying to look for someone with more knowledge to "validate" or to "guide"
>>>>>>> my points of view.
>>>>>>>
>>>>>>> Going through my most recent attempts of having something concrete
>>>>>>> for sharing in plain English I realize one mistake I'm committing: I'm
>>>>>>> trying to describe combustion vehicles (Hypermedia Applications) saying
>>>>>>> that petroleum exists (Semantic Intelligence).
>>>>>>>
>>>>>>> As long as my post are going I've just got a stack of (incoherent)
>>>>>>> "analysis" documents as the result of my work. And I had only those until
>>>>>>> now because I was stuck because of the previously mentioned mistake (ah,
>>>>>>> and because of my Bipolar Disease maniac episodes...).
>>>>>>>
>>>>>>> So, I should try to describe applications instead and see how and
>>>>>>> where fuel should burn properly inside a motion vehicle to generate
>>>>>>> traction. Every semicolon I write is updated into my GitHub repository, so,
>>>>>>> sorry if you browse that "scrapbook" and you don't find anything even
>>>>>>> intelligible.
>>>>>>>
>>>>>>>
>>>>>>> https://github.com/snxama/scrapbook/blob/master/Application.docx?raw=true
>>>>>>>
>>>>>>> There is a brief technical description work in progress document.
>>>>>>> For now it just a list of statements about a potential backend. Session and
>>>>>>> Inteteraction layers are not specified.
>>>>>>>
>>>>>>> https://github.com/snxama/scrapbook/blob/master/Index2.docx?raw=true
>>>>>>>
>>>>>>> Best Regards,
>>>>>>> Sebastian Samaruga:
>>>>>>> http://snxama.blogspot.com
>>>>>>>
>>>>>>>
<https://www.avast.com/sig-email?utm_medium=email&utm_source=link&utm_campaign=sig-email&utm_content=webmail&utm_term=icon>
Virus-free.
www.avast.com
<https://www.avast.com/sig-email?utm_medium=email&utm_source=link&utm_campaign=sig-email&utm_content=webmail&utm_term=link>
<#DAB4FAD8-2DD7-40BB-A1B8-4E2AA1F9FDF2>