Sphinn Home » Other Searching
Avatar Posted By: Eish 58 days ago
Topic Type: News Story (Jump to http://searchengineland.com)
Category: Other Searching
After nearly two years in the making -- and plenty of hype -- Powerset has finally rolled out a "natural language" search engine. It's not a Google killer. It's barely a business model right now. But at least it's something the world can finally play with, and under the hood, there's lots of potential.
3 Comments     

Comments

from demib 57 days ago #
Votes: 1 | Vote:
+ -

I don't see much new in what Powerset is doing. Some years ago I worked as VP og Technology at the Danish "Natural Language Computing" company Ankiro. OK, Danny, I know you don't like the term, but thats what we did :)

What Ankiro did 5-6 years ago - and today do even better - seems a lot more refined and powerfull than what Powerset have to show. Really, what they do looks like some of the very early stages of pre-studies to what Ankiro do. I am not impressed at all!

So far natural language search (in lack of a better word) has proven to work very well on "small" sets of data - such as a Wiki, a single website or other minor collections of documents - not the entire web. The main reason is scalability. None of the technologies for natural language I've seen so far does scale to the numbers global web search requires. And even if it did, I am not sure natural language search will help so much.

What I've experienced natural search do beter than traditional search is to find information when its hard to find any - where the datasets are limited and where you may not know the right words to find it. This is perfect for enterprise search where topics are often described using one "style" of writing and one limited set of words, that may not be what you - as a searcher, have in mind. There is a few other advantages of natural language search but they also do not apply much to web search.
 
In other words, natural language search (so far) helps expand your searches and find matches you otherwise would not find. It makes the "reach" element of search higher but it does not neccesarily make the "precission" element any better. In fact, if you feed a natural language search engine with the billions of unorganised documents on the web the results will (for now) be worse. Higher reach (we don't need that with 20 billion documents and millions of results in each SERP!) and worse precission ... Hmmm, not likely to kill any major engine any day soon :)

from mightyfunk 56 days ago #
Votes: 0 | Vote:
+ -

It says that its technology reads and comprehends each word on a page.

First let's get this straight - It doesn't comprehend anything. That's wishful thinking and marketing. It looks at verbs or certain keywords, flags them as important, references through synonyms, then proceeds to lump them under one category.

It's a smart way to do things, but it's not comprehension. Comprehension would imply artificial intelligence whereas this system follows a set pattern of rules and doesn't 'think' on its own.

In no ways am I trying to put this effort down - it's a step in the right direction. But you have to be careful how you weigh these words.

The second problem I see here is that they choose Wikipedia as an example. I suppose Wikipedia itself is a good site as an example, but it's far from perfect as I've discussed here:
http://www.mightyfunk.com/2008/05/wikipedia-equals-fail-death-to-the-open-encyclopedia/

from dannysullivan 56 days ago #
Votes: 0 | Vote:
+ -

Mightyfunk, good point. I guess I was saying that it understands the words compared to a regular search engine and focused more on that than the contrast between how a human comprehends.


Log in to comment or register here.

Top 10 Most Sphunn
In What's New

  1. 10 Emergency SEO Tips For When Thin... (21)
  2. Positioning yourself using SEO and ... (20)
  3. Importance of Documenting Your SEO ... (20)
  4. PPC Advertising: Art or Science? (20)
  5. 5 ways to improve rankings through ... (19)
  6. How to Switch Your User-Agent to Go... (18)
  7. Fixing duplicate content in SEO sit... (18)
  8. Google passes second link’s anchor ... (17)
  9. Top 10 URL Design Failures of Famou... (17)
  10. Enterprise SEO and the art of the p... (16)

Latest Comments

Search Marketing Expo

Save the date for:
• SMX Local & Mobile - San Francisco, CA (July 24-25) See the agenda, and register now!
• SMX Sao Paolo - Brazil - (Aug. 7-8)
• SMX China - September 23 & 24, 2008
• SMX Stockholm - September 23 & 24, 2008
• SMX East - NYC - (Oct. 6-8) Registration is now open.
• SMX London - November 4 & 5, 2008