Linguists are the new black

Semantic Web Trending in 2008

If Web design and development were fashion, linguists would be the new black in 2008. Beyond the confines of traditional Web vehicles, the need for smart applications is staking its’ claim. Leading this push, are the fledgling efforts of several Web applications, now touting something akin to good old Sci-Fi artificial intelligence.

In a sense, we’re seeing the beginnings of a quasi-Marxist ai, finding more and more space on the web. I’m guessing we’ll see an explosion (poof) in semantic web applications in 2008, even if only from behind the private beta curtain. A fair number of semantic applications are well into development and a few have even made the break into the public domain. But are we ready?

Understanding Communication

The key to the semantics in the Web applications I refer to, is meaning. The theory of signs, as put by the Vienna Circle, can be broken down into 3 parts:

  • Semantics: The relation between signs and the things they refer to.
  • Syntactics: The relation of signs to each other in formal structures.
  • Pragmatics: The relation of signs to their impacts on those who use them.

Web design inherently and irrevocably depends on all three of these disciplines. Where as traditional web applications, and programming for that matter, looks to apply meaning through signs, semantic Web applications attempt to extract meaning from signs. This is normally the part of semiotics left up to us, the humans, the “thinkers,” if you will.

The Algorithm Killed The Cat

Now, the game is to take some of the burden off us over-worked, poor slobs, and let the applications do some of our thinking for us. I kid, but the fact is, I’m all for it. The smarter the team, right? And from where I’m sitting, the Web is most definitely part of my team. You may remember a rather plaintive ad campaign, started by Ask.com sometime in the spring of 2007. The attempt, on behalf of Ask.com, was to distinguish itself from other search engines, by identifying it’s unique algorithm as the source for better results. In a sense, they are crafting the beginnings of a semantic search engine that understands both what the user queries and the available pickings, to return the best matches. This goes beyond a simple or even complex keyword search engine, such as Google. Does it work? Beats me.

the algorithm killed jeeves

The real question is, what was Crispin, Porter + Bogusky, the agency behind this campaign, thinking? The ad campaign slings slogans from “The Algorithm Killed Jeeves” to “The Algorithm Constantly Finds Jesus.” As the Jeeves version adorned billboards in San Francisco, it only struck me as a kind of apology from Ask.com for sucking. In any case, what does this tell the layman? And more importantly, how do you explain a technology that is supposed to think for you?

Thinking so you don’t have to.

Perhaps up and coming semantic Web applications can learn something from our friend Ze, and tout slogans modeled around his famous sign off for the Show, “..thinking so you don’t have to.”

Thinking may be a stretch, but understanding is the challenge (at least from a developer stand point). Semantic applications open the field of Web development to new territory. One of which, being linguistics.

Deriving contextual meaning from text and images is most definitely a task that can produce results nearing that of intelligence and thought (at least on the surface, when done extremely well). To get these results, requires thorough understanding of language, syntax, grammar and how all these constructs are broken. So, it’s not surprising that we’re seeing an increase of linguists on development teams. And maybe this isn’t such an unnatural thing.

As an example, one of San Francisco’s very own semantic application start ups, Powerset, has been working on a natural language search engine for some time now. The idea is that meaning and knowledge are extracted by Powerset, as apposed to keyword based queries. The results reflect something closer to a database of knowledge, rather than a listing of what’s available.

powerset

Powerset is currently in private testing still, but I have a feeling we will be seeing this type of application making a huge impact this year.

Comments

  • Mark says on

    I look forward to natural language and intelligent search engines for one simple reason: horse porn. My portfolio website, http://www.markhogan.com, constantly gets hits from people searching google for “horse porn”. While it is not entirely incorrect, as one of my projects deals with horses and porn, if the search were more intelligent it would realize that it wasn’t what the user was looking for.

  • lauren scime says on

    @mark but how do you know that the user who is searching for “horse porn” wouldn’t find some pleasure in stumbling upon your architecture thesis? Poor search results lend themselves to serendipity. And as my 3rd grade librarian used to say, serendipity is the backbone of education. I’m just sayin’…

  • Mark says on

    My third grade librarian used to say “horse porn is the backbone of education” but that’s probably why I have “issues”.

  • Jeremy Anderson says on

    Dear Valued Commenters,

    I believe you have illustrated quite nicely, why there is such a need for linguists.

    If you mention hp one more time, I will have to educate you!

    I get enough spam, thank you.

Leave a Reply

Submit Comment