KnowledgeUser

A million URLs served!

So what! What good is a million URLs? More is not better!

And why do they keep doing it? Everybody out there that clicks Next through even to page 10 of your search results please raise your mousing hand, and keep it there.

Why not bring back the best results, the results that are thematically weighted to be as relevant as possible to your original query, and then present them in such a way that allows you to learn more about the results than the query you actually put in? There are really two challenges facing the modern Internet searcher:

1. creating a great query that brings back really relevant results

2. having the information presented in such a way that it tells you more about the query results than you would learn by looking at a list of URLs

And probably the biggest obstacle to all of this is TRUST. How can I trust the results I'm getting back? Well, we have the problem now only now you get back one million results. I believe the trust factor can be overcome with better search queries. However, most search engines only allow you to input ten search words into your query which significantly limits your ability to narrow the number of returned URLs.

What if you could use an entire document as your query? You already trust the information that is in the document in question so that addresses part of the trust issue. If the document can be turned into a complex query and send to the search engine we can be sure to get a narrow yet highly relevant result set back. Of the results returned we can select the top ones and generate a Knowledge Interface(TM) that brings insight into the results that have been returned.

Now, you've got something to work with! I'm from Missouri so show me you say!

In the following example, I'm going to use the Cirilab Discovery information triage utility that walks the talk!

1. Right click on any text document and select Cirilab Discovery.

 

2. The document will be sent to the Cirilab Knowledge Generation Engine(TM) and a thematic Knowledge Signature(TM) will be produced. The Knowledge Signature(TM) will then be sent to a search engine. Currently, Google, Yahoo, MSN, and Gigablast are supported search engines.

3. Results will be gathered and individually assessed for their Thematic Resonance(TM) with the original document used to create the query. Only those with the highest relevance ranking will be kept for classification and aggregation.

4. All of the selected results, the ones with the highest Thematic Resonance(TM) to the original query document are aggregated into a Knowledge Map(TM) of the most relevant results.

 

Notice that just under 10 highly relevant results are returned. We could stop there but that would be no different than looking at the first page of the search engine results. If search is going to serve the searcher the results need to provide insight into what was returned. The Document Navigator within a Cirilab Knowledge Map does just that.

As the results were returned a Knowledge Signature(TM) was created for each one. The Knowledge Signatures(TM) are compared for relevance and the highest ones are aggregated into a thematic Knowledge Map(TM). This is where the insight is provided. Themes that were found within the results are presented to the searcher within the Knowledge Map Document Navigator and they provide two benefits:

1. the hierarchical theme structures acts as a auto search facility by allowing the user to further narrow the results to just those that meet a particular thematic vein

2. it introduces themes that were discovered in the returned URLs that the searcher may never have thought of looking for in the first place (that's the Discovery part) and upon finding them may leader the searcher to new and exciting areas of investigation

You can explore the results of the above exercise at:

Cirilab Discovery Results using US Constitution as Query Input

You can see the Knowledge Signature of the original document that was used as the input query.

Cirilab Knowledge Signature of US Constitution

Less search results but more relevant ones are what the search engines should be striving for. And then present the results in such a way that provides insight into what has been returned.

Arnold Villeneuve

   

Diagramming Anniversary

This week CS Odessa celebrates the Anniversary of its flagship diagramming application ConceptDraw VI. The company marks the occasion with the quiz with three copies of ConceptDraw VI Std give away. The quiz lasts from 15 to 22 of November.

To participate visit

http://www.conceptdraw.com/products/time_limited_offer.php

 

Google Sitemaps and Cirilab Knowledge Maps

Are you being indexed Sir/Miss?

I return in a round about way to my previous blog on The Dark Lords of MetaTag Juicing (some of them are my sincere friends like SEO Guru David Jonah). It seems that Google is still bent on putting them out of business. The question remains "how is Google going to evaluate content in the absence of MetaTags as all trust in MetaTags seems to be gone", and rightfully so.

Obviously, Google employees cannot look at every Internet page so some other mechanism seems to be required. That mechanism seems to be "sitemaps". For the webmasters that are not paying attention you are well advised to see the writing on the wall. You see, Google does not care about your MetaTages because they don't believe what you are telling them about the value of their content. Moreover, they have decided to implement a system whereby they (Google) decides the value of your content. This is an important fact to understand because it is a significant shift in the "index me" paradigm.

If you don't see it yet you'll be sorry soon. The paradigm shift is that WebMasters no longer get the option of stating what their content is about but rather that the almighty Google will now decide what your content is about. Google is obviously tired of people abusing MetaTags by telling them that web pages with content about Apples are actually about Oranges and have decided to take action in a serious way to address the issue.

The Emerging Index Paradigm and the SITEMAP!

For those of you that have not stared to look into the emerging concept of a SITEMAP I recommend strongly that you start to look into it in a very serious way because it is the future as Google defines it in the area of Internet search.

There are two sides to the SITEMAP methodology that I can see from my perspective:

1. Tell Google what URLs on your site that you want them to index via a SITEMAP. 

2. Google uses the SITEMAP to crawl the site for the URLs that you specify. It then uses a proprietary and secretive approach to determining the ranking of the content on your website and you have no say in the matter.

Well actually you do, if you know how to use work their system of website and webpage ranking to elevate your search results position. It's all about creating interlinked content and internal website content that refers to trusted external website content. The question is: How do I take all of the content I have now and easily turn it into a "Content Zone(TM)" that Google is willing to rank at a higher level.

That's where Cirilab's Knowledge Maps come into play.  Cirilab's Knowledge Maps are Google Sitemap aware. More importantly, Knolwedge Map content is interlinked because the content is in fact inter-relative within the Content Zone.

There is another key to the success of this approach with Google Sitemaps but you'll need to figure that one out on your own!