New term suggestion: webpage

[Asiyah Yu Lin]

Dear IAO developers,

I wonder how much you will agree to create a term "webpage".

IAO: Webpage definition: A document that written in HTML (hypertext markup language) and is translated by a Web browser.  [definition resource: http://www.techterms.com/definition/webpage]

IAO: Webpage equivalent to https://schema.org/WebPage

is_a document (http://purl.obolibrary.org/obo/IAO_0000310)

Thing foaf:homepae IAO:webpage

webpage is_about Thing

(http://xmlns.com/foaf/spec/#term_homepage)

Please let me know your position.

Thanks,

Asiyah

[Bill Duncan]

Hi Asiyah,

I think this is a good start. Perhaps some other things that might need to be made clear are:

1. The document is electronic, as opposed to printed on a piece of paper. Perhaps you could use the term "html document" for this.

2. The html document is a specified input to the web browser.

3. Perhaps it might  be necessary to distinguish between the content displayed by the web browser after the html has been rendered, and the html document itself.

Your thought?

-- Bill

--
--
informatio...@googlegroups.com
To change settings, visit
http://groups.google.com/group/information-ontology

---
You received this message because you are subscribed to the Google Groups "information-ontology" group.
To unsubscribe from this group and stop receiving emails from it, send an email to information-onto...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[Bill Hogan]

agree with #3.

re: #1: I can print out the html.  Could argue whether paper concretization is still technically a "Web page".  But it seems Cambridge-changish to say it's only a Web page when concretized one way, and not another.

Also #3 interacting with #1: can definitely print rendered page and have them look nearly identical (if not identical for a one-page, rendered page).

Bill

[Bill Duncan]

Yes, a web page can be printed out. So, I suppose you need to be clear if you mean those electronic documents that can be accessed using the internet (via a URL).

[Amanda Hicks]

Hi all,

I agree with Bill Duncan that #3 is important. Werner Ceusters and Shahid Manzoor have a nice paper about referent tracking and webpages. If you haven't seen it already you can find it here. http://www.referent-tracking.com/_RTU/papers/InterOntologyCeusters2009.pdf

I think that following the All/Some rule for definitions, it is good to define a web browser as something that has an html document as a input, but I'm not sure if the web page requires a web browser to process the html in all cases. Certainly web browsers are how people usually interact with webpages, but if a web developer is interacting with the html with a text editor, would we say they are editing the web page? or are they editing something else?

best,

Amanda

Amanda Hicks, PhD

Assistant Professor, Health Outcomes & Policy

College of Medicine

University of Florida

CTRB 3226

PO Box 100219

2004 Mowry Road

Gainesville, FL  32610

p: (352) 294-5932

f: (352) 273-8703

aeh...@ufl.edu

[Jonathan A Rees]

Taking potshots as usual...

I think the BFO ideal is that all the classes that get ontology terms should correspond to universals. Universals do not depend on particulars. "The web" is a particular - it is a thing that exists on this planet in this moment in history. Therefore one could argue "web page" should not be a class.

But maybe there is some universal that is similar. Would a web-like thing on another planet also have "web pages"? Perhaps its URLs would conflict with ours, making the URL-to-page association contingent. How would you decide whether something is a web page?

I don't understand the utility of excluding PDF and all the other formats that exist on the web (or webs). - that is not a complaint, mainly a question.  If you really mean to restrict to HTML, maybe have a class called 'HTML document' (although this is itself tricky since the definition of 'HTML' is a moving target, and therefore contingent, not universal).

[Jonathan A Rees]

The nature of web pages is a notoriously thorny question. Do you really mean a single HTML document? For most modern web sites the HTML is useless by itself, you also need a ton of auxiliary CSS, javascript, and image files, and usually access to a server for continued service. You can't meaningfully print the page without these.

Here's an article about this: http://www.ariadne.ac.uk/issue65/thompson-hs

"What do URLs refer to" was the subject of a long-running argument within the W3C Technical Argument Group that never got resolved and was ultimately dropped (http://www.w3.org/wiki/HttpRange14Webography). The conversation ranged widely over issues like mutability, format variants, customization (e.g. cookies), and the nature of meaning and reference. The question became the butt of jokes and ultimately I think everyone decided the answer didn't matter, if indeed the question was meaningful at all.

I see that the schema.org definition of webpage is "a web page".  That kind of definition would not be very BFO- or IAO-like. Often a term can be useful even in the absence of a definition, as in this case, and that's fine for schema.org. But ontologies are supposed to have definitions, i.e. criteria you can apply to a thing to decide whether that thing is a member of the given class.

I would suggest starting, as with other IAO term requests, with use cases. Then induce definitions from there, following the BFO foundation that IAO seems to be committed to, and then decide what the class should be called.

[Asiyah Yu Lin]

Thank you all for the insightful response.

A use case would be, a gene entity refers to its NCBI gene webpage, or to its HGNC page...

Thanks,

Asiyah

[Bill Duncan]

Hi Asiyah,

It sounds like a 'webpage' in your use case could refer to a couple of things:

1. The display of the information as it presented in the web browser.

2. The information that is accessed via the URL.

It sounds (to me) like you are using #2. In this case, a 'webpage' is probably an aggregate of a number of things, such as the HTML to present the content, the database that holds data about the gene, and the code that accesses the data and translates it into HTML that is to be rendered in the browser. If this is the case, then perhaps you could find a different term than 'webpage'?

-- Bill

[Asiyah Yu Lin]

Thanks, Bill.

What's your suggestion for another term?

[Bill Duncan]

I'm afraid I'm not that creative :(

Let me think about it ...

You could also just be REALLY clear about how you are using the term 'webpage'. This might work for the particular project you are working on.

[Asiyah Yu Lin]

Do you suggest that I may use something like 'gene information webpage display'?

[Bill Duncan]

That might work. Perhaps you could have a class 'webpage display' with 'gene information webpage display' as a subclass. I also thought the term 'rendered webpage' might work. I assume that you are using these terms to represent what is displayed by the web browser ... right?

[Yu Lin]

Correct, Bill.

The rendered webpage was my another thought.

Thanks,

Asian

Sent from my iPad