Here’s a question for the abstract thinking group who frequents the blog:
There are three web sites and three blogs with no apparent connection between them. There are no links between them. They’re found in far corners of the web.
Now, let’s say a web of things emerges, and it’s revealed that the previously disconnected sites are all connected by a common object, in this case a particular book. What do YOU want to do with this?
{ 2 trackbacks }
{ 10 comments… read them below or add one }
Im seeing a visual thesaurus execution…..
Well, right away I want to know more specifically how each is connected to the thing. Is one a review of the book? Is one a fan network for the book? Is one The book? Depending on what I need or want from the concept of that book at this time, one or more of these things may be relatively important to me.
For instance, if I’ve only just heard about the thing, I might want someone’s opinion or analysis of it, so the review, or blog, might be more important then. But if it’s something I already know a great deal about, or enough, anyway, I might be looking to actually find the real thing, out here in the wild. So then a review or fanbase would be irrelevant to me.
Information has no importance other than that which we assign to it at the moment. That may not be popular, but there it is. Another way of putting it is this: For information to be useful it must be personal, relevant and pertinent.
I may decide, after looking at website #3 that it is an undesirable site, full of only links and drive-by infections. I would want to be able to exclude that site from my list of that thing’s connections. Similarly, I would want some way of marking a superior site, and maybe even rank sites publicly and by category (tagging).
On a larger scale, it would be nice to be able to search for something and have categorized results, publicly ranked and organized by people who actually cared about it, ie., people who used it themselves.
There are 3 people out here in the wild. They don’t know each other and will most certainly never be consciously aware of each other’s existence. Person #1 tries out a site about a thing, finds links to some dubious drug instead, and marks it “Undesirable” in his widget. Person #2 tries a site and finds a brilliant review of the thing, and marks it superior. Person #3 hears about the thing, searches with his widget and finds the results, tagged by the others.
Based on their rankings he skips over the link page and goes straight to the brilliant review. They’ve never met, but their experiences have intersected and value has been created. Not just value in the dollar sense, but value in the human sense.
By the way, something funny happened to me on the way over here today… I was clicking around and found Chris’ Nested Universe site and wanted to tell you guys about it… then I realized you already knew….
Nice site.
@Leigh,
A visual thesaurus where you “page” through objects and discover sites that review, sites that discuss, sites that rank, etc. ?
If that’s the idea I’m imaging an implementation similar to the “cover-flow” execution in iTunes where you page through objects until you get one you want to find more about and then you access sites that reference it.
A comment about this is that things like this could exist without a web of things. The question is really how to capitalize on the network that emerges.
Jon, great points and nice thoughts.
“categorized results, publicly ranked and organized by people who actually cared about it” … sounds a little like Leigh’s thoughts in an earlier post, except it’s organized by people who actually cared about it and people you actually care about.
People who care about it and people you actually care about… when it comes to discovering an item which is more important to you?
“They’ve never met, but their experiences have intersected and value has been created”
Last.fm sort of surfaces these relationships through their neighbor radio when it comes to music experiences. They’re not “friendships” in the social network sense, but they’re useful nevertheless.
I actually should have introduced you and Chris, I suspect you’d like one another’s content and thoughts. Chris meet Jon, Jon meet Chris.
In both Leigh’s thesaurus example and Jon’s efficient search description the implementation focuses on finding information about an object that is already known by the individual.
What about other scenarios?
From the point of view from one of the individuals who owns one of the content points that created contributed to the creation of that “web of things” I may want to know more NOT about the item that created the relationship, but more about the owner of the content point.
For example, if that book happens to be a 9/11 conspiracy book, maybe the bloggers that posted about it have either very strong opinions for or against the theory. Then, from that I can find other blogs (content points) that speak to my interest in a 9/11 conspiracy….if I did have that interest.
How to execute that relationship obviously requires more thought…and it’s bed time for me.
…”finding information about an object that is already known by the individual.”
Not necessarily, though. perhaps i’ve just heard someone mention “semantic web” for the first time. I might have no knowledge other than those two words. I’d like to be able to find relevant sites on concepts as well as physical objects.
So in a search through my widget I’d be looking at results and would notice superior ratings for say…… Alex Iskold, or ReadWrite Web, or AdaptivBlue. Being able to sort though nested, rated data that could be trusted would lead me to find Alex, RWW, and this blog, each of which would eventually lead me to many other places and other things (or concepts) I might not have thought of yet.
I like Jon’s point that the relative importance/usefulness of the respective instances of the Thing will depend on each person’s relationship to that thing. Adding this layer of personalization information about “me” makes the underlying web of things exponentially more powerful.
Alex,
You are describing a people and data network expressed as graph. The key question though is this: Will this Data Graph be referencable across data spaces via object identifiers (URIs)? Can I point to data in your data space from mine and vice versa? This is what the Linked Data meme is all about in relation to the Semantic Web.
The Semantic Web has always been about People and Data networks, but we clearly need to draw a line of distinction between:
- Semantic Web Inside solutions with not data referencability outside the application / solution silo
- Semantic Web inside and outside solutions where data is referencable across silos (as esposed by the Linked Data community).
To conclude, you can have a network of things, but how open and accessible is this network? Bearing in mind it’s part of open infrastructure provided by the Web.
I applaud the efforts here to exploit the semantic web. I have been researching and developing collaborative systems for over 30 years. These are exciting times. Usage of the semantic web have been far slower in coming that I have expected. I suppose it is because answering the issue of how we can generalize the notion of information usage is a question without an easy answer.
Usefulness is the value of information whether it is used for survival or entertainment. Usage is a dynamic thing. It is action, not static knowledge.
Thus we need a model of the user, individually and collectively tied to our model of the world such that usage has meaning.
Here is a brainstormed mind map outline of one organization of such a model.
categorization - usage, association, grouping, meaning, influence, effect
redundancies
indicators
contra-indicators
categorization contexts by perspective
objective
absolute
evidence
intrinsic properties
relative contexts
agreement, reinforcement evidence
purpose or effect
subjective
collective contexts
ratings /rankings / orderings
popularity / agreement / reinforcement
individual contexts
context trusts
ratings /rankings / orderings
reinforcement
implied relations
similar patterns
implied issues
unexpected association missing or present
human actions (individual or group)
expected
actual
potential patterns
agent actions (individual or group)
explicit rules
reinforced rules
random rules
While the above is half baked, it suggests the complexity of the semantic network that is focused on doing and not just being. It has a dynamical meta ontological nature that butters the toast ;-)
I hope I have not confused the issue. My point is that it is a tough but important question. Answering this issue is key to making the semantic web truly useful.
Jim