Semantical linking

[kgun]

Professor Erik Wilde and Dr. David Lowe have coauthored an inspiring and thoughtprovoking book:

XPath, XLink, XPointer, and XML: A Practical Guide to Web Hyperlinking and Transclusion Addison Wesley ISBN: 0201703440 that inspired med to write this short post and pose the following questions:

1. How do you define semantical linking?
2. Do you think some SE's are getting better at identifying semantical links?
3. Do any of you use the HTML DOM innerHTML Property that sets and returns text of links?
4. If yes to 3? Do you think, since JS is used, that it can negatively affect your ranking, since semantics may be lost?
5. Got an idea: An advanced SeBot can transform a malformed HTML document and related documents (that it links to) to well-formed X(HT)ML documents, pick the XML Information Set and analyze content, context and link semantics in a much more advanced way than done in the pure HTML web.
6. This raises serious questions about the permanency of links and / or content. Personally I have observed the following: Even large important companies and providers delete or move links. Cool URL's don't change.
7. As a consequence of point 6, by deleting links (URI's), changing content or moving links (URI's) without redirection, you may penalize yourslef by being regarded as a less serious content provider.
8. Perma links. Is that an important label? Should you inform in a subheading how permanent your links (URI's) and subject are? Eg by issuing statements like:
   - This content is not final and may be changed.
   - This content is final and the link (URI) will not be deleted in 24 months.
   - This content and this link (URI) is final and will be permanent as long as the site lives.

[wige]

I can only speak to a couple of these points...
3. I have used the innerHTML property on several of my sites, but not for modifying links in real time. Because this can be applied to any HTML tag, it provides the viewing portion of Ajax, to display output and change options for the user.
4. It seems to me that search engines continue to look at web pages as static objects that do not change after the page is loaded. I have seen behavior that indicates they try to figure out the effects of changes to the page, but since this data is very easy to manipulate and not frequently used, I think the SEs would only be willing to provide a very limited capability to parse this.
6. Documents on the web need to have the flexibility to change... that is kind of the point. So
7. I think this means that we need smarter servers, rather than smart documents. The HTTP and HTML specs both provide incomplete mechanisms for handling the transient state of the Internet. For example, there is no way for a server to tell a client a document does not exist (404 means the document is unavailable, which is why search engines keep coming back, to see if the document has become available - if you return a 410, the spiders will stop asking because that says the file existed but was deleted. A method to tell the spider the page never existed in the first place would be extremely useful).
8. This is already addressed in the HTTP and HTML standards, both provide, through the cache control mechanisms, ways to communicate to a client the expected permanence of a document.

Honestly, I think semantic linking will be a pass-by technology. It offers a lot of potential functionality that is not currently used, and that there is a limited need for right now, and most of what this enables a developer to do can be accomplished, in many ways more effectively, through other more commonplace mechanisms.

[kgun]

Very few comments on this topic that is the heart of the WWW. I found this

Fortune Interactive's Research Sheds a Light on How Search Engines Determine Rankings

interesting article today. Fortune Interactive has a tool, SEMLogic

Some related links:
Download 'The Forrester Wave™: Enterprise Search Q2 2008' - VivÃ*simo, Inc.

Google Universal Search: 2008 Edition

I have started to use Clusty (owned by VivÃ*simo, Inc. - Enterprise Search, Federated Search and Clustering ) more in my web search and the results are often, IMO, better than the big three and the clustered hits to the left of the SERP's often categorize better than the Big three.




KW's for Clusty is:

• Meta search.
• Clustering
  
• Comparative ranking.

Related WPW threads:
Semantic web, semantic ...

What is the monetary value of a link?

Experiment:

Compare the SERP's on the big three and Clusty for some KW's and you may note that where the big three give the Wikipedia result Clusty give another result. Am I correct?

[crankydave]

Originally Posted by kgun

Compare the SERP's on the big three and Clusty for some KW's and you may not that where the big three give the Wikipedia result Clusty give another result. Am I correct?

I agree kgun. I have used Clusty for some time and actually like many of their results. I do wish it was a bit more extensive though but not to the point of clutter.

Dave

[deepsand]

I've been using MetaCrawler for years.

It returns both organic and paid results from Google, Yahoo, Windows Live, Ask and others, and indicates which type of listing and from which source(s) each result was returned.

It also provides for searching by type/source (Web, Image, Audio, Video, News, Yellow Pages & White Pages), along with some personalization/preferences.

MetaCrawler - Wikipedia, the free encyclopedia

[kgun]

It seems to me that both of them get a lot of hits from Ask and Clusty get ads from Ask as far as I know.

You can compare hits for both of them on my new search resource page that is submitted for review.

[deepsand]

Originally Posted by kgun

It seems to me that both of them get a lot of hits from Ask and Clusty get ads from Ask as far as I know.

You can compare hits for both of them on my new search resource page that is submitted for review.

Submitted for review where?

[kgun]

http://www.surftoolbar.com/ : : Constructive feedback, please not on markup.

[janeth]

I think it is a very good topic.

* Off-page optimization factors out-weighed any on-page optimization factors.

Nothing new there.

* In-bound link quality was the most important factor across all three engines. However, each engine's optimal range for link quality was different.

Nothing new there

* In determining in-bound link quality, the reputation of the originating web page was more important than the page's relevance to the keyword "laptop."

That would blow away these people that believe the site needs to be on topic with your site.

* In-bound link quantity was the least important factor among off-page factors.

Would explain why site wide links do not help much.

* The most important on-page factor for Google was title-tag keyword density.

I need to look into this a little.

* Web pages that successfully ranked across the engines all had strong values in at least the two most important influential factors for each search engine respectively.

Need to look at this more as well.

[janeth]

In-Bound Link (IBL) Quality . This is a measurement of key elements on the page containing an in-bound link which, in combination, influence the link reputation for the target of the link. This is the only factor that had the same level of relative influence across the search engines and happened to be the most influential in all cases.

Do we know what those key elements are?

[kgun]

Janeth, I am in an extreme hurry. Fine that you participate. A few reflections:

1. Semantic Linking and semantic search (engines) will (is) in my view be an important concept.
2. Google transclusion. A concept you have to learn sooner than later if you do not already know it.
3. You should have participated in this Ignore the misunderstandings that have produced the present world. thread.
4. Listen to this Transclusion: Fixing Electronic Literature video if you have time.
5. Two other important of my sites: W3Toolbar.com: A collection of resources directly or indirectly related to The World Wide Web Consortium (W3C) and Cybertoolbar.com: A collection of popular resources related to the cyberspace. (under construction).