Re: Argumentation Markup

Adam,

making claims URI-addressable is a cool idea, but why use new HTML tags 
and attributes, or a new format?

In the first section, you list a number of standard-compliant 
alternative (e.g. RDFa). How are they not fulfilling your use-cases?

And even if they are suboptimal, the cost-benefit ratio compared to 
extending HTML is likely to be better.

Anyway, that's my 2¢

On 13/01/2021 03:34, Adam Sobieski wrote:
>
> Argumentation Community Group,
>
> Math Working Group,
>
> Semantic Web Interest Group,
>
> Hello. I would like to share some new ideas for discussion.
>
> HTML and MathML content could be wrapped in new markup elements (e.g. 
> <fact> or <claim>) or could be adorned with new attributes (e.g. meta) 
> to provide the means for automated reasoning technologies and/or users 
> to traverse to metadata, provenance, mathematical proof, or 
> argumentation about the contained HTML and MathML content.
>
> These preliminary ideas are available online 
> (https://www.w3.org/community/argumentation/wiki/Argumentation_Markup 
> <https://www.w3.org/community/argumentation/wiki/Argumentation_Markup>) 
> and are also indicated below.
>
>
>     *URI-addressable Facts and Claims*
>
> Some preliminary ideas for representing URI-addressable facts or 
> claims in HTML documents include:
>
> 1.new markup elements (e.g. |<fact id="...">| or |<claim id="...">|)
>
> 2.text fragments (https://wicg.github.io/scroll-to-text-fragment/ 
> <https://wicg.github.io/scroll-to-text-fragment/>)
>
> 3.standardized class 
> <https://html.spec.whatwg.org/multipage/dom.html#classes> names (e.g. 
> |<span id="..." class="w3c-fact">|)
>
> 4.the role <https://www.w3.org/TR/role-attribute/> attribute (e.g. 
> |<span id="..." role="fact">|)
>
> 5.custom elements (e.g. |<fact-span id="...">|)
>
> 6.Web schema (see: http://schema.org <http://schema.org/>)
>
> 7.RDFa or similar technologies
>
> 8.embedding semantics in HTML documents via |<script>| elements to 
> indicate which URI-addressable document elements are facts or claims
>
>
>     *New Markup Elements*
>
> With new markup for facts or claims (e.g. |<fact id="...">| or |<claim 
> id="...">|), an “argument Web” could be composed where URI-addressable 
> facts and claims in HTML documents could each reference their 
> supporting and/or opposing argumentation, such argumentation 
> referencing other facts or claims by URI.
>
> *<fact*id="..."href="..."*>*HTML and MathML content*</fact>*
> *<claim*id="..."href="..."*>*HTML and MathML content*</claim>*
>
> In the above example, the |href| attribute is for referencing a 
> resource which could include metadata, provenance data, mathematical 
> proof, or argumentation supporting and/or opposing the fact or claim. 
> The referenced resource could be text, XML, HTML, RDF, AIF, or another 
> format.
>
>
>       Scenario: Inline Argumentation
>
> For scenarios where it is desired to include metadata, provenance 
> data, mathematical proof, or argumentation in the same HTML document 
> which contains a fact or claim, one can use |<script>| elements with 
> ID’s to provide inline resources.
>
> <*html*>
> <*head*>
> <*script*id="inline-proof-123"type="...">...</*script*>
> </*head*>
> <*body*>
> <*fact*id="fact-123"href="#inline-proof-123">HTML and MathML 
> content</*fact*>
> </*body*>
> </*html*>
>
> The content in such |<script>| elements can refer to URI-addressable 
> facts or claims in containing documents as well as in other documents 
> or resources.
>
>
>     *A New Attribute*
>
> Understanding the meaning of the |href| attribute on a |<fact>| or 
> |<claim>| element, we could rename that |href| attribute to, for 
> example, |meta| which would mean that it could adorn existing HTML 
> elements.
>
> <*html*>
> <*head*>
> <*script*id="inline-proof-123"type="...">...</*script*>
> <*script*id="inline-proof-124"type="...">...</*script*>
> </*head*>
> <*body*>
> <*span*id="fact-123"role="fact"meta="#inline-proof-123">HTML and 
> MathML content</*span*>
> <*div*id="fact-124"role="fact"meta="#inline-proof-124">HTML and MathML 
> content</*div*>
> </*body*>
> </*html*>
>
>
>     *A New Format for Metadata, Provenance, Mathematical Proof, and
>     Argumentation*
>
> A new format, model, or ontology could be of use for the 
> machine-utilizable resources referenced by URI-addressable facts or 
> claims, resources containing metadata, provenance, mathematical proof, 
> or argumentation.
>
> <*html*>
> <*head*>
> <*script*id="inline-proof-123"type="application/new-format">...</*script*>
> <*script*id="inline-proof-124"type="application/new-format">...</*script*>
> </*head*>
> <*body*>
> <*span*id="fact-123"role="fact"meta="#inline-proof-123">HTML and 
> MathML content</*span*>
> <*div*id="fact-124"role="fact"meta="#inline-proof-124">HTML and MathML 
> content</*div*>
> <*span*id="fact-125"role="fact"meta="https://www.example.org/automated-reasoning.php?fact=125">HTML 
> and MathML content</*span*>
> </*body*>
> </*html*>
>
>
>     *Embedding Metadata, Provenance, Mathematical Proof, and
>     Argumentation in HTML Documents*
>
> Alternatively, resources could be (X)HTML and contain structured data 
> for metadata, provenance, mathematical proof, or argumentation about a 
> fact or claim. In this scenario, in addition to being 
> machine-utilizable, resources would be human-readable.
>
> <*html*>
> <*body*>
> <*span*id="fact-123"role="fact"meta="https://www.example.org/automated-reasoning/123.xhtml">HTML 
> and MathML content</*span*>
> <*div*id="fact-124"role="fact"meta="https://www.example.org/automated-reasoning/124.xhtml">HTML 
> and MathML content</*div*>
> <*span*id="fact-125"role="fact"meta="https://www.example.org/automated-reasoning/125.xhtml">HTML 
> and MathML content</*span*>
> </*body*>
> </*html*>
>
>
>       The Link Element
>
> For (X)HTML documents which are about a URI-addressable fact or claim, 
> e.g. containing metadata, provenance, mathematical proof, or 
> argumentation, we could use the |<link>| element, resembling:
>
> <*html*>
> <*head*>
> <*link*rel="about"href="http://www.example.org/document.xhtml#fact-123"/>
> </*head*>
> <*body*>
>     ...
> </*body*>
> </*html*>
>
> Best regards,
>
> Adam Sobieski
>
> http://www.phoster.com/ <http://www.phoster.com/>
>

Received on Friday, 15 January 2021 09:30:26 UTC