(continued imaginary conversation between Angela Tesoro and Timothy Falconer, now sitting in her board room with her financial advisors)
Angela: Everyone, this is Tim Falconer. He’s been helping me understand this Semantic Web stuff in the last few days. I’m now convinced there’s a need for the technology, but I still haven’t heard the most important part: will this make us money?
Timothy: I’m convinced it can, particularly given the ideas in the prospectus you gave me. Your startup company essentially wants to create a worldwide technical talent search system, like Monster.com, only decentralized. I showed you FOAF yesterday so you could see what this might look like.
Angela: So our startup could use FOAF?
Timothy: They could build on it, but the prospectus indicates they’d rather make their own specialized metadata vocabulary, or “ontology.” It’s actually not very difficult to make the ontology, or even the system that uses it. The tricky part, the risky part, is convincing lots of people to annotate their resumes using the new ontology. Until then, they’ll just have some nice useless software and no customers.
Angela: So they gotta get buy-in before it’ll take off. This sounds like the bad press we’ve been reading, that the Semantic Web’s a nice idea, but people won’t take to it, like the metric system in America.
Timothy: That’s certainly the elephant in the middle of the room, but I think you’ll have a much easier time achieving buy-in because you’re targeting a technical audience. Software and web developers are more likely to mark up their resumes with RDF. Given our still struggling technical economy, I think they’ll also be motatived to do it, particularly once they see it can give them an edge in finding work.
Angela: What about the real customers, the companies paying to use the search system. You run a software company. Would you pay for it?
Timothy: Absolutely. I’d love to use a system like this. With it, I’d be able to search for developers who have five or more years of Java experience, who understand object modeling, and design patterns, and refactoring, who have experience with particular technologies like EJB or Swing, who have a reasonable hourly rate, and who have experience working on their own. Being able to do a search like this would make my life much easier.
Angela: But how could you be sure they weren’t lying about their experience? What’s that thing you say, “Garbage In, Garbage Out”?
Timothy: Well, there’s an important piece of the Semantic Web puzzle that addresses that issue, the so-called “web of trust.” The idea is that people can vouch for the accuracy of other people’s metadata. This “trust” can be encoded in the system in a secure fashion, which means you can then do searches that only bring up trustworthy results.
Angela: But can’t the vouching people lie too?
Timothy: They could, but it wouldn’t matter. The measure of trust could be grounded by someone I already knew and trusted. Just like in real life, if I trust someone, who trusts someone, who trusts someone, I’m more likely to find a good worker. It’s the same as using personal recommendations, but it’s systematized, so it saves time. I think it’s a great idea.
Angela: Okay, well you’ve given us a lot to think about. Thanks again for your advice, and your patience. I’ll let you know if we decide to invest in this Semantic Web startup.
Timothy: It was my pleasure. Call anytime. Good luck!
Posted in Angela Talk, semantic web | Comments Off
(continued imaginary conversation between Angela Tesoro and Timothy Falconer, sitting with a laptop at their local WiFi equipped bookshop)
Angela: Since yesterday, I’ve been doing “page source” on a bunch of web pages like you showed me. Sure enough, lots of them have these meta keywords. I even gave some thought to our keywords and had our webmaster change them after looking at some competitor websites.
Timothy: Meta keywords are just a small example of where we’re headed with the Semantic Web. They’re baby steps at best. To do better, we need to annotate stuff using shared metadata vocabularies.
Angela: Explain.
Timothy: Let me show you. Open up your home page again.
Angela: I really gotta update this thing. I never look at it.
Timothy: Let’s see. There’s a picture of you, and your name, and your phone, and your email, and a bio which talks about the companies you’re connected with, where you went to school, and so on. Now, if a search engine, or some other software, wanted to learn what college you went to, it wouldn’t be able to do it easily because it’d have to interpret your bio, which is much easier for a human to do than a computer program. Imagine if someone wanted to find all women CEOs who went to small liberal arts colleges. With the current Web, this would be a very difficult task.
Angela: But not if my home page said, “This thing is a college.”
Timothy: Right, if we annotated your homepage with some metadata, using a shared vocabulary, the search engines could be smarter about it.
Angela: Show me.
Timothy: Okay, here’s an example of some real Semantic Web metadata, written in a format called RDF, using a shared vocabulary called FOAF. It describes someone named Mark Pilgrim. Skip past the “bc” stuff and see if you can tell me what it means.
Angela: Uh, okay. There’s his name, and he’s an ISTP. That’s cool. That looks like his email address, and his homepage, and his work homepage. He works in DC and he went to school in Indiana. Looks like there’s some images or something and he’s interested in Macs and Zen Buddhism. It shows what projects he’s worked on, and … now this is cool … it lists the people he knows.
Timothy: Yeah, that’s a big part of it, in fact it’s why they call it FOAF, which means Friend-Of-A-Friend. It’s a way of linking people together so computers can do lots of nifty stuff.
Angela: So you’re saying I can ask for all the people who went to Earlham College who like Macintoshes and Zen Buddhism?
Timothy: Something like that. You can also browse interpersonal relationships, like board members of large corporations. Click the green circles to see who they’re connected to. The fatter the cat, the more companies they’re connected to.
Angela: Oh, this is just too cool. Let me play for a while.
Timothy: I gotta run anyway. See you tomorrow.
Posted in Angela Talk, semantic web | Comments Off
(continued imaginary conversation between Angela Tesoro and Timothy Falconer, now sitting in her office in front of her computer)
Angela: Yesterday at lunch you were telling me the world needs more metadata, and that metadata is a kind of one-off description of “real” data. What I don’t understand is why this is new. Isn’t just about everything we do with computers related to metadata? My friend Julie uses annotation in Word all the time. And isn’t every form we fill out on the web like this? I type in “Angela Tesoro” in the “Name” field. Isn’t “Name” meta to “Angela”?
Timothy: I can see you’ve been thinking about this.
Angela: (smiles) Yeah, well, they want a lot of money.
Timothy: You’re right mostly. “Data about data” has been around forever. The new part is how we’re gonna use metadata on the Web. Before we get into that, though, let’s step back a bit and talk about the web as it is now. Go ahead and open up your personal home page.
Angela: Okay, but it’s not that great. There.
Timothy: Great picture. Now this page has already been marked up, but it’s not the Semantic Web kind of markup. Choose “View / Page Source” from the menu bar. There. That’s HTML, which means the HyperText Markup Language. Those things between the angle brackets are called tags. They’re data about data, but here it’s used mostly to tell the web browser how things should look. See those <b> tags? That makes your company name bold. Those <i> tags make your email italic.
Angela: What about <title>?
Timothy: Very good. That’s not just about looks. That’s more “meta”, more “semantic”. It tells the browser the title of the page, just like a subject on an email.
Angela: What about these <meta> ones? Are these Semantic Web things?
Timothy: Your quick! In a way, they are, or at least they were the only game in town when this page was made. One of them gives a general description of the website. The other gives some keywords appropriate to the website, like “women” and “fitness” and “workout”. These are used by the search engines to help people find you.
Angela: Wow, I didn’t know this was in here. So your saying if I type in “fitness” in a search engine, my website will show up because of these keywords.
Timothy: Well, it’d show up, but you’d probably have to do a lot of digging. You see, search engines aren’t very good at drawing fine distinctions. They just look at the raw text. If you were researching backyard whirlpools, you’d have to wade through a lot of stuff about the company named Whirlpool. There’s currently no way to say “ignore the company”, because search engines don’t even know that Whirlpool is the name of a company. They don’t know that in some cases, Whirlpool means “name of company”.
Angela: Again with the semantics. I think I see where this is going. I hate to do this, but unfortunately, I gotta run. More tomorrow?
Timothy: See you then.
Posted in Angela Talk, semantic web | Comments Off
(continued imaginary conversation between Angela Tesoro, billionaire founder of a fictious women’s health club franchise, and Timothy Falconer, advocate for the Semantic Web)
Angela: Okay, now that we’ve ordered lunch, tell me about the Semantic Web. What is it? Why should I invest in it?
Timothy: Let’s start with some definitions. There’s basically two pieces to the Semantic Web puzzle: 1) annotating content with metadata, and 2) doing stuff with that metadata.
Angela: I’m already lost. You sound like those guys from yesterday.
Timothy: (smiling) I’ll explain. By “content”, I mean anything you can make with a computer or put into a computer: Word documents, emails, digital photos, spreadsheets, and so on. “Annotating” is a ten dollar word that means adding descriptive information, like a teacher who marks up a term paper. “Metadata” is simply the markups, the red scribblings on the term paper. You start with content (the term paper), then you annotate it (scribble on it), which gives you metadata (scribblings).
Angela: A little clearer. So why is this good?
Timothy: Well, first understand that there’s nothing new about annotation or metadata. You do it every day when you send an email. You write your message (content), then you add the subject and the recipient (metadata). This extra stuff isn’t really part of the message. It’s about your message, data about data, hence metadata.
Angela: I see. (laughs) Why can’t you people talk like normal people?
Timothy: That’d be too easy. Besides, it’s how we keep getting the big bucks, like lawyers.
Angela: Ain’t that the truth. So I’m guessing your gonna tell me the world needs more metadata.
Timothy: Yes. We need lots and lots of it. See, the Web’s this big scattered mess of content, and we need better ways to find what we’re looking for, like the Dewey Decimal System does for libraries. We need metadata to help us figure out what it all means.
Angela: So you’re saying the Semantic Web is sorta like the current Web, except everybody’s added subject lines to all the web pages.
Timothy: Kinda. (Food arrives) Let’s eat, and talk more later.
Posted in Angela Talk, semantic web | Comments Off
(what follows is an imaginary conversation between Angela Tesoro, billionaire founder of a fictious women’s health club franchise, and Timothy Falconer, advocate for the Semantic Web)
Angela: Hi Tim, it’s Angela. Got a minute?
Timothy: Hey Angela. Sure, what’s up?
Angela: This morning I was approached by a startup company that’s looking to make and sell some computer software. They’re convinced they’re gonna make millions of course, and they need an angel investor, which is why they met with me. Their ideas sound pretty good, but you know me. I’m not the most computer literate person, which is why I’m calling you.
Timothy: What’s their idea?
Angela: It’s probably too complicated to go into right now. I’ll have my assistant send you the prospectus. From what I understand though, the whole thing hinges on the success of something called the “Semantic Web.” We did some Google searches and found some pretty negative commentary about it, though. My financial advisor thinks we should pass because too many people are saying it’s some kind of pipe dream. Remember how much I lost in the dot-com bust? We’re still pretty gun-shy over here, but I figured I’d call you first to see what you thought. What do you think of the Semantic Web?
Timothy: I think it’s the real deal. There’s definitely money to be made if you play your cards right. What did they tell you about it?
Angela: A lotta stuff about something called ‘metadata’, and how it’ll let everyone make better websites. Most of what they said went over my head. They said it’s the vision of the guy who invented the Internet.
Timothy: You mean the guy who invented the World Wide Web.
Angela: (laughs) If you say so. What’s the difference?
Timothy: The Web is just one of many things you can do on the Internet. There’s also email, and instant messenger, and a bunch of other things. They all use the Internet, which is like the plumbing that holds it all together. The Internet’s been around a lot longer than the Web. Most people first learned about the Internet when the Web became popular, so it’s a common mistake to confuse the two. And yes, the Semantic Web is an improvement on the first Web, led by the guy who invented the first Web, Tim Berners-Lee.
Angela: Well it sounds like I’m talking to the right person. Can we have lunch tomorrow? I need someone to teach me about this Semantic Web.
Timothy: Sure thing. See you tomorrow.
Posted in Angela Talk, semantic web | Comments Off
to describe his vision of the Semantic Web (used with permission)
"Tidepool" and "Storymill" are trademarks of Immuexa Corporation.
Website design copyright © 2003-2004 by Immuexa.
