The semantic future
I have been reading a lot about the semantic web of late and am wondering
what it will hold for the future of online and more specifically for us
(Ixd/IA/UE etc etc).
For instance it appears to me that if the semantic web becomes a reality
then data becomes the most important thing online and not web sites. For
instance If I want to buy a TV online in the future I will just enter my
criteria into an application and it will return all matches from across the
web. I won't need to go to Amazon, ebay, PC World etc, The data will come to
me. This is obviously already happening online especially with content in
the form of RSS. I don't need to go to 50 content providers every day my RSS
feeds brings them to me. This trend in many ways is a good thing because it
levels the playing field somewhat, for instance smaller content producers or
ecommerce operations will be able to more readily compete with the big names
as Search results, marketing budgets, eye-balls and traffic become less
important and querying of standardised data becomes the norm.
So my question is, what do you think that the semantic web will bring? Will
it ultimately bring about the death of the web page? And if so where does
that leave us? Or will it only enhance our designs and sites, allowing us to
bring more choice to the consumer?
Sorry for the vagueness of the question but just interested if anyone has
any thoughts or must reads in this area?
Get the feed
Comments
It seems to me that the death of the web page would coincide with the death
of marketing...and I know that's not going to happen. I think semantic
content retrieval is a good thing, since it seems to be a standardization
around a more efficient way to retrieve one particular kind of information -
stuff you want to know and that you have asked for. You asked for it either
by performing a query (as you mention, with TVs), or by subscribing to
topical feeds where content from many sites is aggregated. On the other
hand, that still leaves a big area open - the type of content that orients
someone around a particular subject or brand or company. In a purely
semantic world where there are no web sites, just aggregators and search
engines (assuming more intelligent ones than we have now), how will I
discover information about a new subject about which I know nothing? I
could perform a search and get back isolated random fragments, news
clippings, products and other bits of information, but I'm still going to
want to go to a "home" for an expert on the subject. If I look for that
kind of information outside of the semantic world, someone else is going to
want to be there waiting with a brand, a message, a mission statement, and
(more likely than not) a ton of branding graphics and images.
The semantic world works well for the just-in-time, I-need-it-now type of
information, but no one is ever going to post their "About Our Company"
message to an RSS feed, and if I'm looking for a job I'm going to want to
find that somewhere. Wikipedia probably won't be the best place for it.
Perhaps some form of Wikis will become that "master home for all About Us
pages". I dunno...
Good question =]
PS: The thoughts above are the result of Tuesday morning musings, and do
not accurately reflect the thoughts of the left side of my brain.
Bryan
http://www.bryanminihan.com
-----Original Message-----
From: discuss-bounces at lists.interactiondesigners.com
[mailto:discuss-bounces at lists.interactiondesigners.com] On Behalf Of Dan
So my question is, what do you think that the semantic web will bring? Will
it ultimately bring about the death of the web page? And if so where does
that leave us? Or will it only enhance our designs and sites, allowing us to
bring more choice to the consumer?
Sorry for the vagueness of the question but just interested if anyone has
any thoughts or must reads in this area?
I think the rise of the semantic web will coincide with the shift to
hand-held devices as the primary data source. Eventually your hand-held
devices will not only be able to speak to you, but they'll know when you're
in the mood to hear what they have to say. Wireless providers will form
partnerships with commercial interests (which is what will *really *force
current marketing departments to evolve, as the emphasis becomes about
delivering a good experience...and not just a lot of *talk *about a good
experience) and everyday life will become very interesting indeed.
I'm betting there will always be a need for a web page, for the reasons you
mentioned and also because designers still need a canvas upon which to
create. It's fascinating to be at such a point in history where we have to
design for both ends of the spectrum, not just in terms of technology, but
for audiences as well.
-G
On Jan 15, 2008, at 6:15 AM, Dan wrote:
> I have been reading a lot about the semantic web of late and am
> wondering
> what it will hold for the future of online and more specifically
> for us
> (Ixd/IA/UE etc etc).
Oh this will be a fun thread... 8^)
I wouldn't worry or sweat about the "semantic web" too much. (This
coming from a web standards kind of guy to boot.) The semantic web as
a means of writing markup is only useful to the point it's not, which
in a structured layout world, is quite often. I write my markup as
semantic as I can until I hit those barriers. Then I do whatever I
need to in order to get the job done.
Easy example: H# Tags. The concept of numbered headers that also
designate indentation and nesting only ever made sense in a linear
content world, like a novel or technical manual. It makes no sense in
the newspaper world, where headlines are designed to work so that
stories can start or continue anywhere on the page, and are stylized
for that purpose. In the structured layout world of the web, I often
use H# tags as a means or simply identifying different classes of
styles to use for things like headlines, where their intention was to
create organizational logic and nesting of meaning. Using H# tags to
denote only the type of headline one wants is not "semantic" but oh
well... That's how I'm going to use it since there is no other clean
way for me as a designer to create headlines for non-linear content.
Another example: The Float property in CSS. It was designed to work
to allow simple images or illustrations to be nested inside body copy
with the text flowing around it. It wasn't really designed for
anything else. However, since in HTML there is still no way to create
a legitimate column or layout structure, designers like to use Float
to forces DIVs to create the column structure. However, given the
nature of Float and given how a computer reads in a linear stream
(machine language, PostScript, all of these do the same thing) you
have to put markup in a certain order for the float effects to do
what you want to create a structured grid. This sometimes means the
markup has to be written out of order, which prevents it from being
easily human readable, which if my memory serves me, is one of the
goals of being semantic markup; being human readable. (It is possible
given certain conditions to have the markup still be in order to make
the effect work, but when its not, you simple change the order.)
For a good example of these issues at work, check out the new
CNN.com. I think the redesign was pretty good at making CNN far more
readable and approachable than the old one. It follows a nice grid
structure, the typography is spec'd nicely and gets better when you
dive further in to the site and it's a lot cleaner to look at. On the
whole, I think they did a really nice job with it. If you view its
page source, you'll also note that they did a really good job of
keeping most of the markup pretty clean. But is their markup
"semantic." Sort of, but not quite. A lot of that has to do with
creating structure to make the visual presentation work, and using
tricks when doing so.
So in that sense, the semantic web is a nice goal to have, but don't
get too caught up in trying to make it a 100% reality as its not
really 100% possible except for simple and linear content on the web,
which is becoming more and more rare. HTML+CSS would have to be
written from scratch to solve structured layout for the semantic web
to truly happen 100%, and getting that happen is probably going to be
as hard as getting people to stop using double spaces after periods
in the writing. 8^)
However, like CNN, use the lessons taught in the semantic web to help
you create cleaner code and markup. Cleaner code which is a result
from web standards and attempts at being "semantic" is more
lightweight to write, manage and pass over the tubes, and will always
be better than other methods.
--
Andrei Herasimchuk
Principal, Involution Studios
innovating the digital world
e. andrei at involutionstudios.com
c. +1 408 306 6422
I would argue that data already is the most important thing online, not
websites. Web "sites" are the containers for our data. Seems to me that the
most innovative things happening and yet to happen in the online space are
in the form of platform based systems that deliver data to the user in a
consumable form. Though I agree with Bryan that there will always be a place
for the traditional "site" a la "About Our Company" which just illustrates
that sometimes we need data that is nicely packaged and wrapped in a pretty
bow.
Celeste
Dan
"For instance it appears to me that if the semantic web becomes a reality
then data becomes the most important thing online and not web sites. "
> ________________________________________________________________
> *Come to IxDA Interaction08 | Savannah*
> February 8-10, 2008 in Savannah, GA, USA
> Register today: http://interaction08.ixda.org/
>
> ________________________________________________________________
> Welcome to the Interaction Design Association (IxDA)!
> To post to this list ....... discuss at ixda.org
> Unsubscribe ................ http://www.ixda.org/unsubscribe
> List Guidelines ............ http://www.ixda.org/guidelines
> List Help .................. http://www.ixda.org/help
>
I agree with Gloria, though I'm a little wary about a device that
knows my mood! Her points are good ones, though, as is the question.
The Semantic Web applies to Web pages, too, in a slightly different
but related context. Semantic structure is already a big benefit to
web designers who use it, but will be an even bigger benefit in the
future. The code I write today is ready to transition to portable
devices because it is structured and because the presentational
aspects are separated. Subtract the presentational data, substitute a
suitable presentational schema for the latest semantically aware
whizbang device, and you're good to go forward into yet another
brave new world.
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Posted from the new ixda.org
http://www.ixda.org/discuss?post=24551
On Jan 15, 2008, at 6:15 AM, Dan wrote:
> I have been reading a lot about the semantic web of late and am
> wondering
> what it will hold for the future of online and more specifically
> for us
> (Ixd/IA/UE etc etc).
Oh this will be a fun thread... 8^)
I wouldn't worry or sweat about the "semantic web" too much. (This
coming from a web standards kind of guy to boot.) The semantic web as
a means of writing markup is only useful to the point it's not, which
in a structured layout world, is quite often. I write my markup as
semantic as I can until I hit those barriers. Then I do whatever I
need to in order to get the job done.
Easy example: H# Tags. The concept of numbered headers that also
designate indentation and nesting only ever made sense in a linear
content world, like a novel or technical manual. It makes no sense in
the newspaper world, where headlines are designed to work so that
stories can start or continue anywhere on the page, and are stylized
for that purpose. In the structured layout world of the web, I often
use H# tags as a means or simply identifying different classes of
styles to use for things like headlines, where their intention was to
create organizational logic and nesting of meaning. Using H# tags to
denote only the type of headline one wants is not "semantic" but oh
well... That's how I'm going to use it since there is no other clean
way for me as a designer to create headlines for non-linear content.
Another example: The Float property in CSS. It was designed to work
to allow simple images or illustrations to be nested inside body copy
with the text flowing around it. It wasn't really designed for
anything else. However, since in HTML there is still no way to create
a legitimate column or layout structure, designers like to use Float
to forces DIVs to create the column structure. However, given the
nature of Float and given how a computer reads in a linear stream
(machine language, PostScript, all of these do the same thing) you
have to put markup in a certain order for the float effects to do
what you want to create a structured grid. This sometimes means the
markup has to be written out of order, which prevents it from being
easily human readable, which if my memory serves me, is one of the
goals of being semantic markup; being human readable. (It is possible
given certain conditions to have the markup still be in order to make
the effect work, but when its not, you simple change the order.)
For a good example of these issues at work, check out the new
CNN.com. I think the redesign was pretty good at making CNN far more
readable and approachable than the old one. It follows a nice grid
structure, the typography is spec'd nicely and gets better when you
dive further in to the site and it's a lot cleaner to look at. On the
whole, I think they did a really nice job with it. If you view its
page source, you'll also note that they did a really good job of
keeping most of the markup pretty clean. But is their markup
"semantic." Sort of, but not quite. A lot of that has to do with
creating structure to make the visual presentation work, and using
tricks when doing so.
So in that sense, the semantic web is a nice goal to have, but don't
get too caught up in trying to make it a 100% reality as its not
really 100% possible except for simple and linear content on the web,
which is becoming more and more rare. HTML+CSS would have to be
written from scratch to solve structured layout for the semantic web
to truly happen 100%, and getting that happen is probably going to be
as hard as getting people to stop using double spaces after periods
in the writing. 8^)
However, like CNN, use the lessons taught in the semantic web to help
you create cleaner code and markup. Cleaner code which is a result
from web standards and attempts at being "semantic" is more
lightweight to write, manage and pass over the tubes, and will always
be better than other methods.
--
Andrei Herasimchuk
Principal, Involution Studios
innovating the digital world
e. andrei at involutionstudios.com
c. +1 408 306 6422
I got a message privately that made a point I might be confusing the
semantic web with something else. (Probably something closer to what
most designers know as "semantic markup.") It's a good point, so I
figure I'd write my response to the list in case I am.
----------
> I think you must be thinking of something else, the issues you
> raise have little/nothing to do with the foundation of the semantic
> web: OWL representations of the concepts your site's content is
> presenting to humans.
I'm probably conflating issues and making certain leaps too early in
what I wrote, but for the semantic web to occur in my understanding
of the term, the data representation has to work to do more than
simply present the data. It has to define he meaning of the data so
more can be done with it.
From: http://www.sciam.com/article.cfm?
id=00048144-10D2-1C70-84A9809EC588EF21&page=2
"The Semantic Web is not a separate Web but an extension of the
current one, in which information is given well-defined meaning,
better enabling computers and people to work in cooperation. The
first steps in weaving the Semantic Web into the structure of the
existing Web are already under way. In the near future, these
developments will usher in significant new functionality as machines
become much better able to process and "understand" the data that
they merely display at present."
To give "information well-defined meaning" you are effectively
tagging it, and to do so separate from the visual presentation can be
a problem if you also use the markup to define its meaning for the
purpose of letting the computer figure out what to do with it.
For example:
<h1>This is my title</h1>
and
<p class="mainheader">This is my title</p>
Are two very different things. One is using the tag to define the
type of information and could be used by the computer to do all sorts
of behaviors with it, but requires all data of the type to be
visually the same. The other is using CSS to simply style it, and
dropping the class removes the style from the object itself. In order
for the semantic web to work, web markup should follow the first
method, not the second. (In my understanding.) And yet, to do complex
content presentation in a web browser, example #2 is actually easier
to control and allows the designer to make the content work better
for the specific design case in question.
So what we end up with is something more like this to try and solve
both issues at the same time:
<div class="article">
<h1>This is my title</h1>
<p>This is my story ... </p>
</div>
And then use
div.article h1 {
style: properties go here;
}
In CSS to declare H1 styles and presentation strictly confined to
content with article classes, while also giving the computer a means
to use H1 for something useful outside of the presentation because it
has special things it can do with content defined as H1. It's also
nice because we can now use things like JQuery to create interaction
using the class selectors in ways that I think model the intention of
the "semantic web," and to use JQuery effectively, it requires
cleaner markup that should be a "semantic" as possible to keep the
Javascript workable.
This kind of "semantic markup" approach to markup works to a degree,
but it quickly gets messy once the content gets more complex and the
presentation more complex. When that mess occurs, one usually has to
toss out "semantic markup" to get the presentation and interaction
correct. And in tossing out semantic markup, one also is impacting
the semantic web.
Again, in my understanding. I could easily be conflating these two
issues when that's not necessary.
I'll admit I know little about the backend data model representation
that is being proposed for the semantic web or if there's going to be
some new means of translating that meaning into usable markup for
browser presentation (outside of all the XML tools that exist today),
but if it's anything like what is happening currently, it's a
hodgepodge and mashed up set of data mixing things semantically with
things that are not. And when that happens, getting the job done
wins, not getting the job done in a way that some future set of new
features might do some useful things.
So yes, I'm probably conflating and getting very far ahead of the
issue and making all sorts of leaps that I passed over in my initial
email, along with leaps that aren't correct, if I'm not confusing the
issue in the first place. 8^) If I am, my apologies.
--
Andrei Herasimchuk
Principal, Involution Studios
innovating the digital world
e. andrei at involutionstudios.com
c. +1 408 306 6422
Thanks for your responses, I find this a really interesting topic and your
thoughts have been equally interesting. Here are a couple more thoughts of
mine below:
Bryan...
>on the other hand, that still leaves a big area open - the type of content
that orients
>someone around a particular subject or brand or company. In a purely
>semantic world where there are no web sites, just aggregators and search
>engines (assuming more intelligent ones than we have now), how will I
>discover information about a new subject about which I know nothing?
Good point, there will always be more information needed regarding a
transaction online than the product details alone. For instance delivery
details, company details, refund policy etc which may need to be contained
within a conventional site. However if the vision of the semantic web is
fulfilled it should make it easy to discover information on a new subject
which I know nothing, by having the ability to pull in all related
information for your needs, rather than you having to move from page to page
to find that information.
Gloria...
>I think the rise of the semantic web will coincide with the shift to
hand-held devices as the primary data source
I agree, I think the mobile has the potential to play a big part in the
semantic web. It really goes back to Bryan's point that the Semantic web
would be great for the type of 'I need to know now' information, and the
mobile means this can happen anywhere, anytime.
>I would argue that data already is the most important thing online, not
websites. Web "sites" are the containers for our data.
I agree to a point, obviously the most important thing online is to find
answers to our questions, Products for our needs etc. However, is it not
Information contained within sites that is the most important thing
currently, rather than raw data which can be sliced and diced as we require.
Going back to my example earlier for instance, if I currently want to find
out who is the most reputable retailer of TV's I can search the web for
'information' and reviews. Maybe in the future I just query the web, get the
'data' and turn it into information tailored for my needs. (sorry that is a
little vague, this is probably an indication in it's self that I am wrong
;-) )
In general, my question regarding the death of the web page was trying to
stimulate discussion, and I am sure there will always be a need for it,
after all not everything we do online revolves around retail. We socialise,
play, write and contribute in discussion etc. But I do wonder if the
Semantic web becomes a reality, will companies invest less in transactional
websites and focus more upon standardising data so that it can be passed to
a semantic agent (application) which will do the work for them.
On Jan 15, 2008 8:57 PM, Andrei Herasimchuk <andrei at involutionstudios.com>
wrote:
> I got a message privately that made a point I might be confusing the
> semantic web with something else. (Probably something closer to what
> most designers know as "semantic markup.") It's a good point, so I
> figure I'd write my response to the list in case I am.
>
> ----------
>
> > I think you must be thinking of something else, the issues you
> > raise have little/nothing to do with the foundation of the semantic
> > web: OWL representations of the concepts your site's content is
> > presenting to humans.
>
>
> I'm probably conflating issues and making certain leaps too early in
> what I wrote, but for the semantic web to occur in my understanding
> of the term, the data representation has to work to do more than
> simply present the data. It has to define he meaning of the data so
> more can be done with it.
>
> From: http://www.sciam.com/article.cfm?
> id=00048144-10D2-1C70-84A9809EC588EF21&page=2
>
> "The Semantic Web is not a separate Web but an extension of the
> current one, in which information is given well-defined meaning,
> better enabling computers and people to work in cooperation. The
> first steps in weaving the Semantic Web into the structure of the
> existing Web are already under way. In the near future, these
> developments will usher in significant new functionality as machines
> become much better able to process and "understand" the data that
> they merely display at present."
>
> To give "information well-defined meaning" you are effectively
> tagging it, and to do so separate from the visual presentation can be
> a problem if you also use the markup to define its meaning for the
> purpose of letting the computer figure out what to do with it.
>
> For example:
>
> <h1>This is my title</h1>
>
> and
>
> <p class="mainheader">This is my title</p>
>
> Are two very different things. One is using the tag to define the
> type of information and could be used by the computer to do all sorts
> of behaviors with it, but requires all data of the type to be
> visually the same. The other is using CSS to simply style it, and
> dropping the class removes the style from the object itself. In order
> for the semantic web to work, web markup should follow the first
> method, not the second. (In my understanding.) And yet, to do complex
> content presentation in a web browser, example #2 is actually easier
> to control and allows the designer to make the content work better
> for the specific design case in question.
>
> So what we end up with is something more like this to try and solve
> both issues at the same time:
>
> <div class="article">
> <h1>This is my title</h1>
> <p>This is my story ... </p>
> </div>
>
> And then use
>
> div.article h1 {
> style: properties go here;
> }
>
> In CSS to declare H1 styles and presentation strictly confined to
> content with article classes, while also giving the computer a means
> to use H1 for something useful outside of the presentation because it
> has special things it can do with content defined as H1. It's also
> nice because we can now use things like JQuery to create interaction
> using the class selectors in ways that I think model the intention of
> the "semantic web," and to use JQuery effectively, it requires
> cleaner markup that should be a "semantic" as possible to keep the
> Javascript workable.
>
> This kind of "semantic markup" approach to markup works to a degree,
> but it quickly gets messy once the content gets more complex and the
> presentation more complex. When that mess occurs, one usually has to
> toss out "semantic markup" to get the presentation and interaction
> correct. And in tossing out semantic markup, one also is impacting
> the semantic web.
>
> Again, in my understanding. I could easily be conflating these two
> issues when that's not necessary.
>
> I'll admit I know little about the backend data model representation
> that is being proposed for the semantic web or if there's going to be
> some new means of translating that meaning into usable markup for
> browser presentation (outside of all the XML tools that exist today),
> but if it's anything like what is happening currently, it's a
> hodgepodge and mashed up set of data mixing things semantically with
> things that are not. And when that happens, getting the job done
> wins, not getting the job done in a way that some future set of new
> features might do some useful things.
>
> So yes, I'm probably conflating and getting very far ahead of the
> issue and making all sorts of leaps that I passed over in my initial
> email, along with leaps that aren't correct, if I'm not confusing the
> issue in the first place. 8^) If I am, my apologies.
>
> --
> Andrei Herasimchuk
>
> Principal, Involution Studios
> innovating the digital world
>
> e. andrei at involutionstudios.com
> c. +1 408 306 6422
>
>
> ________________________________________________________________
> *Come to IxDA Interaction08 | Savannah*
> February 8-10, 2008 in Savannah, GA, USA
> Register today: http://interaction08.ixda.org/
>
> ________________________________________________________________
> Welcome to the Interaction Design Association (IxDA)!
> To post to this list ....... discuss at ixda.org
> Unsubscribe ................ http://www.ixda.org/unsubscribe
> List Guidelines ............ http://www.ixda.org/guidelines
> List Help .................. http://www.ixda.org/help
>
Hi Dan,
something that was not really mentioned so far is that there are still
two major problems associated with the deployment of the Semantic Web.
First of all, there is the issue of trust. Not all information can be
equally trusted, so your traditional Semantic Web "shopping scenario"
would fall apart unless there is a way to make sure your results are
clean from rogue retailers, non-existent offers and all kinds of
related spam. Looking at current websites like Wikipedia, there is
still a lot to be desired in terms of trust and verifiability of
information, despite their best efforts to sort these problems out. I
see this as a big challenge to which interaction design can definitely
contribute.
Second, there is no way of making a great ontology of everything, that
everyone will agree with (ontology is a data model that represents a
set of concepts within a domain and the relationships between those
concepts). This is even for small domains like shopping: when I moved
to another country a couple of years ago I was initially quite puzzled
with the way some things were categorised in UK online shopping
websites, which was not quite the same taxonomy that I had formed in
my brain while growing up in another part of the world. Again, it's a
big challenge how to reconcile these different views of the world if
you'd like to produce some unified set of results.
An interesting (although maybe a bit outdated) article that mentions
the above problems in more detail can be found at:
http://www.techreview.com/Infotech/12640/
Some more sceptic opinions at:
http://www.shirky.com/writings/semantic_syllogism.html
http://www.oreillynet.com/xml/blog/2006/06/the_7_flaws_of_the_semantic_we.html
Regards,
Alex
On Jan 16, 2008 11:19 AM, Dan <dgwillia at googlemail.com> wrote:
> Thanks for your responses, I find this a really interesting topic and your
> thoughts have been equally interesting. Here are a couple more thoughts of
> mine below:
>
> Bryan...
> >on the other hand, that still leaves a big area open - the type of content
> that orients
> >someone around a particular subject or brand or company. In a purely
> >semantic world where there are no web sites, just aggregators and search
> >engines (assuming more intelligent ones than we have now), how will I
> >discover information about a new subject about which I know nothing?
>
> Good point, there will always be more information needed regarding a
> transaction online than the product details alone. For instance delivery
> details, company details, refund policy etc which may need to be contained
> within a conventional site. However if the vision of the semantic web is
> fulfilled it should make it easy to discover information on a new subject
> which I know nothing, by having the ability to pull in all related
> information for your needs, rather than you having to move from page to page
> to find that information.
>
> Gloria...
> >I think the rise of the semantic web will coincide with the shift to
> hand-held devices as the primary data source
>
> I agree, I think the mobile has the potential to play a big part in the
> semantic web. It really goes back to Bryan's point that the Semantic web
> would be great for the type of 'I need to know now' information, and the
> mobile means this can happen anywhere, anytime.
>
>
> >I would argue that data already is the most important thing online, not
> websites. Web "sites" are the containers for our data.
>
> I agree to a point, obviously the most important thing online is to find
> answers to our questions, Products for our needs etc. However, is it not
> Information contained within sites that is the most important thing
> currently, rather than raw data which can be sliced and diced as we require.
> Going back to my example earlier for instance, if I currently want to find
> out who is the most reputable retailer of TV's I can search the web for
> 'information' and reviews. Maybe in the future I just query the web, get the
> 'data' and turn it into information tailored for my needs. (sorry that is a
> little vague, this is probably an indication in it's self that I am wrong
> ;-) )
>
> In general, my question regarding the death of the web page was trying to
> stimulate discussion, and I am sure there will always be a need for it,
> after all not everything we do online revolves around retail. We socialise,
> play, write and contribute in discussion etc. But I do wonder if the
> Semantic web becomes a reality, will companies invest less in transactional
> websites and focus more upon standardising data so that it can be passed to
> a semantic agent (application) which will do the work for them.
>
>
>
>
> On Jan 15, 2008 8:57 PM, Andrei Herasimchuk <andrei at involutionstudios.com>
> wrote:
>
>
> > I got a message privately that made a point I might be confusing the
> > semantic web with something else. (Probably something closer to what
> > most designers know as "semantic markup.") It's a good point, so I
> > figure I'd write my response to the list in case I am.
> >
> > ----------
> >
> > > I think you must be thinking of something else, the issues you
> > > raise have little/nothing to do with the foundation of the semantic
> > > web: OWL representations of the concepts your site's content is
> > > presenting to humans.
> >
> >
> > I'm probably conflating issues and making certain leaps too early in
> > what I wrote, but for the semantic web to occur in my understanding
> > of the term, the data representation has to work to do more than
> > simply present the data. It has to define he meaning of the data so
> > more can be done with it.
> >
> > From: http://www.sciam.com/article.cfm?
> > id=00048144-10D2-1C70-84A9809EC588EF21&page=2
> >
> > "The Semantic Web is not a separate Web but an extension of the
> > current one, in which information is given well-defined meaning,
> > better enabling computers and people to work in cooperation. The
> > first steps in weaving the Semantic Web into the structure of the
> > existing Web are already under way. In the near future, these
> > developments will usher in significant new functionality as machines
> > become much better able to process and "understand" the data that
> > they merely display at present."
> >
> > To give "information well-defined meaning" you are effectively
> > tagging it, and to do so separate from the visual presentation can be
> > a problem if you also use the markup to define its meaning for the
> > purpose of letting the computer figure out what to do with it.
> >
> > For example:
> >
> > <h1>This is my title</h1>
> >
> > and
> >
> > <p class="mainheader">This is my title</p>
> >
> > Are two very different things. One is using the tag to define the
> > type of information and could be used by the computer to do all sorts
> > of behaviors with it, but requires all data of the type to be
> > visually the same. The other is using CSS to simply style it, and
> > dropping the class removes the style from the object itself. In order
> > for the semantic web to work, web markup should follow the first
> > method, not the second. (In my understanding.) And yet, to do complex
> > content presentation in a web browser, example #2 is actually easier
> > to control and allows the designer to make the content work better
> > for the specific design case in question.
> >
> > So what we end up with is something more like this to try and solve
> > both issues at the same time:
> >
> > <div class="article">
> > <h1>This is my title</h1>
> > <p>This is my story ... </p>
> > </div>
> >
> > And then use
> >
> > div.article h1 {
> > style: properties go here;
> > }
> >
> > In CSS to declare H1 styles and presentation strictly confined to
> > content with article classes, while also giving the computer a means
> > to use H1 for something useful outside of the presentation because it
> > has special things it can do with content defined as H1. It's also
> > nice because we can now use things like JQuery to create interaction
> > using the class selectors in ways that I think model the intention of
> > the "semantic web," and to use JQuery effectively, it requires
> > cleaner markup that should be a "semantic" as possible to keep the
> > Javascript workable.
> >
> > This kind of "semantic markup" approach to markup works to a degree,
> > but it quickly gets messy once the content gets more complex and the
> > presentation more complex. When that mess occurs, one usually has to
> > toss out "semantic markup" to get the presentation and interaction
> > correct. And in tossing out semantic markup, one also is impacting
> > the semantic web.
> >
> > Again, in my understanding. I could easily be conflating these two
> > issues when that's not necessary.
> >
> > I'll admit I know little about the backend data model representation
> > that is being proposed for the semantic web or if there's going to be
> > some new means of translating that meaning into usable markup for
> > browser presentation (outside of all the XML tools that exist today),
> > but if it's anything like what is happening currently, it's a
> > hodgepodge and mashed up set of data mixing things semantically with
> > things that are not. And when that happens, getting the job done
> > wins, not getting the job done in a way that some future set of new
> > features might do some useful things.
> >
> > So yes, I'm probably conflating and getting very far ahead of the
> > issue and making all sorts of leaps that I passed over in my initial
> > email, along with leaps that aren't correct, if I'm not confusing the
> > issue in the first place. 8^) If I am, my apologies.
> >
> > --
> > Andrei Herasimchuk
> >
> > Principal, Involution Studios
> > innovating the digital world
> >
> > e. andrei at involutionstudios.com
> > c. +1 408 306 6422
> >
> >
> > ________________________________________________________________
> > *Come to IxDA Interaction08 | Savannah*
> > February 8-10, 2008 in Savannah, GA, USA
> > Register today: http://interaction08.ixda.org/
> >
> > ________________________________________________________________
> > Welcome to the Interaction Design Association (IxDA)!
> > To post to this list ....... discuss at ixda.org
> > Unsubscribe ................ http://www.ixda.org/unsubscribe
> > List Guidelines ............ http://www.ixda.org/guidelines
> > List Help .................. http://www.ixda.org/help
> >
> ________________________________________________________________
> *Come to IxDA Interaction08 | Savannah*
> February 8-10, 2008 in Savannah, GA, USA
> Register today: http://interaction08.ixda.org/
>
> ________________________________________________________________
> Welcome to the Interaction Design Association (IxDA)!
> To post to this list ....... discuss at ixda.org
> Unsubscribe ................ http://www.ixda.org/unsubscribe
> List Guidelines ............ http://www.ixda.org/guidelines
> List Help .................. http://www.ixda.org/help
>
Thanks for the link's I have added these to my Delicious account for later
reading.
In a similar vein to your point, tone of voice and branded content will
become more important as they will increasingly live and breath outside of a
site. This is already important with RSS.
In fact I think this topic is mentioned within the following article:
http://www.digital-web.com/articles/web_2_for_designers/
I wonder if in the future we would have a greater dependence upon the Agents
(apps, borwsers) to decide which are credible sources of information and
which are not.
On Jan 16, 2008 12:08 PM, Alexander Baxevanis <alex.baxevanis at gmail.com>
wrote:
> Hi Dan,
>
> something that was not really mentioned so far is that there are still
> two major problems associated with the deployment of the Semantic Web.
>
> First of all, there is the issue of trust. Not all information can be
> equally trusted, so your traditional Semantic Web "shopping scenario"
> would fall apart unless there is a way to make sure your results are
> clean from rogue retailers, non-existent offers and all kinds of
> related spam. Looking at current websites like Wikipedia, there is
> still a lot to be desired in terms of trust and verifiability of
> information, despite their best efforts to sort these problems out. I
> see this as a big challenge to which interaction design can definitely
> contribute.
>
> Second, there is no way of making a great ontology of everything, that
> everyone will agree with (ontology is a data model that represents a
> set of concepts within a domain and the relationships between those
> concepts). This is even for small domains like shopping: when I moved
> to another country a couple of years ago I was initially quite puzzled
> with the way some things were categorised in UK online shopping
> websites, which was not quite the same taxonomy that I had formed in
> my brain while growing up in another part of the world. Again, it's a
> big challenge how to reconcile these different views of the world if
> you'd like to produce some unified set of results.
>
> An interesting (although maybe a bit outdated) article that mentions
> the above problems in more detail can be found at:
> http://www.techreview.com/Infotech/12640/
>
> Some more sceptic opinions at:
> http://www.shirky.com/writings/semantic_syllogism.html
>
> http://www.oreillynet.com/xml/blog/2006/06/the_7_flaws_of_the_semantic_we.html
>
> Regards,
> Alex
>
> On Jan 16, 2008 11:19 AM, Dan <dgwillia at googlemail.com> wrote:
> > Thanks for your responses, I find this a really interesting topic and
> your
> > thoughts have been equally interesting. Here are a couple more thoughts
> of
> > mine below:
> >
> > Bryan...
> > >on the other hand, that still leaves a big area open - the type of
> content
> > that orients
> > >someone around a particular subject or brand or company. In a purely
> > >semantic world where there are no web sites, just aggregators and
> search
> > >engines (assuming more intelligent ones than we have now), how will I
> > >discover information about a new subject about which I know nothing?
> >
> > Good point, there will always be more information needed regarding a
> > transaction online than the product details alone. For instance delivery
> > details, company details, refund policy etc which may need to be
> contained
> > within a conventional site. However if the vision of the semantic web is
> > fulfilled it should make it easy to discover information on a new
> subject
> > which I know nothing, by having the ability to pull in all related
> > information for your needs, rather than you having to move from page to
> page
> > to find that information.
> >
> > Gloria...
> > >I think the rise of the semantic web will coincide with the shift to
> > hand-held devices as the primary data source
> >
> > I agree, I think the mobile has the potential to play a big part in the
> > semantic web. It really goes back to Bryan's point that the Semantic web
> > would be great for the type of 'I need to know now' information, and the
> > mobile means this can happen anywhere, anytime.
> >
> >
> > >I would argue that data already is the most important thing online, not
> > websites. Web "sites" are the containers for our data.
> >
> > I agree to a point, obviously the most important thing online is to find
> > answers to our questions, Products for our needs etc. However, is it not
> > Information contained within sites that is the most important thing
> > currently, rather than raw data which can be sliced and diced as we
> require.
> > Going back to my example earlier for instance, if I currently want to
> find
> > out who is the most reputable retailer of TV's I can search the web for
> > 'information' and reviews. Maybe in the future I just query the web, get
> the
> > 'data' and turn it into information tailored for my needs. (sorry that
> is a
> > little vague, this is probably an indication in it's self that I am
> wrong
> > ;-) )
> >
> > In general, my question regarding the death of the web page was trying
> to
> > stimulate discussion, and I am sure there will always be a need for it,
> > after all not everything we do online revolves around retail. We
> socialise,
> > play, write and contribute in discussion etc. But I do wonder if the
> > Semantic web becomes a reality, will companies invest less in
> transactional
> > websites and focus more upon standardising data so that it can be passed
> to
> > a semantic agent (application) which will do the work for them.
> >
> >
> >
> >
> > On Jan 15, 2008 8:57 PM, Andrei Herasimchuk <
> andrei at involutionstudios.com>
> > wrote:
> >
> >
> > > I got a message privately that made a point I might be confusing the
> > > semantic web with something else. (Probably something closer to what
> > > most designers know as "semantic markup.") It's a good point, so I
> > > figure I'd write my response to the list in case I am.
> > >
> > > ----------
> > >
> > > > I think you must be thinking of something else, the issues you
> > > > raise have little/nothing to do with the foundation of the semantic
> > > > web: OWL representations of the concepts your site's content is
> > > > presenting to humans.
> > >
> > >
> > > I'm probably conflating issues and making certain leaps too early in
> > > what I wrote, but for the semantic web to occur in my understanding
> > > of the term, the data representation has to work to do more than
> > > simply present the data. It has to define he meaning of the data so
> > > more can be done with it.
> > >
> > > From: http://www.sciam.com/article.cfm?
> > > id=00048144-10D2-1C70-84A9809EC588EF21&page=2
> > >
> > > "The Semantic Web is not a separate Web but an extension of the
> > > current one, in which information is given well-defined meaning,
> > > better enabling computers and people to work in cooperation. The
> > > first steps in weaving the Semantic Web into the structure of the
> > > existing Web are already under way. In the near future, these
> > > developments will usher in significant new functionality as machines
> > > become much better able to process and "understand" the data that
> > > they merely display at present."
> > >
> > > To give "information well-defined meaning" you are effectively
> > > tagging it, and to do so separate from the visual presentation can be
> > > a problem if you also use the markup to define its meaning for the
> > > purpose of letting the computer figure out what to do with it.
> > >
> > > For example:
> > >
> > > <h1>This is my title</h1>
> > >
> > > and
> > >
> > > <p class="mainheader">This is my title</p>
> > >
> > > Are two very different things. One is using the tag to define the
> > > type of information and could be used by the computer to do all sorts
> > > of behaviors with it, but requires all data of the type to be
> > > visually the same. The other is using CSS to simply style it, and
> > > dropping the class removes the style from the object itself. In order
> > > for the semantic web to work, web markup should follow the first
> > > method, not the second. (In my understanding.) And yet, to do complex
> > > content presentation in a web browser, example #2 is actually easier
> > > to control and allows the designer to make the content work better
> > > for the specific design case in question.
> > >
> > > So what we end up with is something more like this to try and solve
> > > both issues at the same time:
> > >
> > > <div class="article">
> > > <h1>This is my title</h1>
> > > <p>This is my story ... </p>
> > > </div>
> > >
> > > And then use
> > >
> > > div.article h1 {
> > > style: properties go here;
> > > }
> > >
> > > In CSS to declare H1 styles and presentation strictly confined to
> > > content with article classes, while also giving the computer a means
> > > to use H1 for something useful outside of the presentation because it
> > > has special things it can do with content defined as H1. It's also
> > > nice because we can now use things like JQuery to create interaction
> > > using the class selectors in ways that I think model the intention of
> > > the "semantic web," and to use JQuery effectively, it requires
> > > cleaner markup that should be a "semantic" as possible to keep the
> > > Javascript workable.
> > >
> > > This kind of "semantic markup" approach to markup works to a degree,
> > > but it quickly gets messy once the content gets more complex and the
> > > presentation more complex. When that mess occurs, one usually has to
> > > toss out "semantic markup" to get the presentation and interaction
> > > correct. And in tossing out semantic markup, one also is impacting
> > > the semantic web.
> > >
> > > Again, in my understanding. I could easily be conflating these two
> > > issues when that's not necessary.
> > >
> > > I'll admit I know little about the backend data model representation
> > > that is being proposed for the semantic web or if there's going to be
> > > some new means of translating that meaning into usable markup for
> > > browser presentation (outside of all the XML tools that exist today),
> > > but if it's anything like what is happening currently, it's a
> > > hodgepodge and mashed up set of data mixing things semantically with
> > > things that are not. And when that happens, getting the job done
> > > wins, not getting the job done in a way that some future set of new
> > > features might do some useful things.
> > >
> > > So yes, I'm probably conflating and getting very far ahead of the
> > > issue and making all sorts of leaps that I passed over in my initial
> > > email, along with leaps that aren't correct, if I'm not confusing the
> > > issue in the first place. 8^) If I am, my apologies.
> > >
> > > --
> > > Andrei Herasimchuk
> > >
> > > Principal, Involution Studios
> > > innovating the digital world
> > >
> > > e. andrei at involutionstudios.com
> > > c. +1 408 306 6422
> > >
> > >
> > > ________________________________________________________________
> > > *Come to IxDA Interaction08 | Savannah*
> > > February 8-10, 2008 in Savannah, GA, USA
> > > Register today: http://interaction08.ixda.org/
> > >
> > > ________________________________________________________________
> > > Welcome to the Interaction Design Association (IxDA)!
> > > To post to this list ....... discuss at ixda.org
> > > Unsubscribe ................ http://www.ixda.org/unsubscribe
> > > List Guidelines ............ http://www.ixda.org/guidelines
> > > List Help .................. http://www.ixda.org/help
> > >
> > ________________________________________________________________
> > *Come to IxDA Interaction08 | Savannah*
> > February 8-10, 2008 in Savannah, GA, USA
> > Register today: http://interaction08.ixda.org/
> >
> > ________________________________________________________________
> > Welcome to the Interaction Design Association (IxDA)!
> > To post to this list ....... discuss at ixda.org
> > Unsubscribe ................ http://www.ixda.org/unsubscribe
> > List Guidelines ............ http://www.ixda.org/guidelines
> > List Help .................. http://www.ixda.org/help
> >
>
I'm afraid I've had a delay in getting this off so some has already been covered...
I understood the SW that Dan originally referred to be the SW that facilitates the application of globally understood meaning to the data. By having this common understanding, one can make then make inferences across the data. Whilst one can apply semantics with mark-up, its very limited if one's heading is 'bills' is that duck bills or what you ask the waiter for (checks).
As to the original question, I agree with Gloria's suggestion that the hand-held will become increasingly important. In fact the Tim Berners-Lee et al key article in Scientific American<http://www.sciam.com/article.cfm?articleID=00048144-10D2-1C70-84A9809EC588EF21>, is about hand-helds in context sensitive situations. He uses an example of organising care for a relative but it could be anything where external relationships (of all kinds) impact on a desired activity.
>From a UI perspective, the SW will hopefully increase the quality of data so e.g. if you search for tables, you can specify that you mean 'furniture' and not a set of 'figures'. This may allow the more elaborate search results presentation e.g. if you only provide few high quality results, you may want to display more data about those results. Patterns of presentation may be developed whereby data from different sources is identified or given prominence.
As Alex mentions, SW apps may require a lot more trust, if your app has made inferences across a load of data and come up with a solution then will you trust it? How will the interface flag what's coming from a trusted source and what isn't?
Incidentally, I stumbled across this recently which appears to have some interesting resources wrt UI and the SW - IPGems<http://www.ipgems.com/swui/>
Ivor
Ivor Tillier
Senior Web Producer
ivor.tillier at oxon.blackwellpublishing.com<mailto:ivor.tillier at oxon.blackwellpublishing.com>
At 11:19 AM +0000 1/16/08, Dan wrote:
>Thanks for your responses, I find this a really interesting topic and your
>thoughts have been equally interesting. Here are a couple more thoughts of
>mine below:
>
<snip>
>In general, my question regarding the death of the web page was trying to
>stimulate discussion, and I am sure there will always be a need for it,
>after all not everything we do online revolves around retail. We socialise,
>play, write and contribute in discussion etc. But I do wonder if the
>Semantic web becomes a reality, will companies invest less in transactional
>websites and focus more upon standardising data so that it can be passed to
>a semantic agent (application) which will do the work for them.
I actually suspect that the reverse may occur in the business sphere
and that marketing requirements -- whether real or perceived -- will
lead to a general situation in which the standardized/semantic web
data is restricted to internal use and the public, transactional
sites are geared toward the rhetorical presentation of the
information.
It will be interesting to see how it plays out. Indeed, it will be
interesting to see *if* it plays out.
Katie
--
----------------
Katie Albers
katie at firstthought.com