Jump to content

something to check on your website


Recommended Posts

>Ease of navigation - a spider can easily follow links within the site - it can access the homepage in one 'jump' from

>any page in the site, links are not JScript links, not using Session ID's

You are confusing how easy it is for the search engine to follow links and find pages from whether that is an indicator of quality of the site. I really don't see why a link back to the homepage is an indicator of quality. To the user, a JavaScript or Session ID link makes no difference to their experience of the site so why should a search engine rank them differently?

>Spiders read all content - if its readable, not difficult to scan a help section, determine the scope of the page >compared to other 'similar pages' from other sites - is it better or worse, then score based on that

These all sound fine, but it is impossible even to determine which are the help pages, let alone determine their "scope". Computers don't understand English, it is just pattern matching.

>Links are easily scamed and SE's need other ways to determine 'quality' - how would you suggest they do it?

Yes links are easily scammed - but scammed links are relatively easy to detect and avoid. There have been a number of approaches, from the vey simple ignore duplicate links from the same IP subnet/subdomain, detection of automatic link patterns, to the most recent being Trust Rank.

It is not a conjecture that links are the primary means used by search engines to determine quality (in the case that links exist!), it is a demonstratable mathematical fact. It's not news either, PageRank *is* the Google patent you referred to. The Google toolbar shows the PageRank which is the query-independent score, which has been shown to be link derived.

If that is not enough, why not listen to what Google have to say?

http://www.google.com/intl/en/technology/

"The heart of our software is PageRank™, a system for ranking web pages developed by our founders Larry Page and Sergey Brin at Stanford University. And while we have dozens of engineers working to improve every aspect of Google on a daily basis, PageRank continues to provide the basis for all of our web search tools.

PageRank Explained

PageRank relies on the uniquely democratic nature of the web by using its vast link structure as an indicator of an individual page's value...."

Link to comment
Share on other sites

>> To the user, a JavaScript or Session ID link makes no difference to their experience of the site so why should a search engine rank them differently?
It certainly does if the user's browser has problems with jScript, if the user is disabled for example and uses a screen reader - if a spider has difficulty in following a link within a site that *may* make several pages of the site unavailable to the search engine and therefore missed by potential searches
It is realtivly easy NOT to use JScript and Session IDs in navigation - in fact it may be the easy way out to use these - better sites will avoid using both - IMO

Sure links have been one of the main factors used by SE's (Google for sure) but that is not the 'be all and end all' of it - recent blog spam and 'contests' based on link spam have caused SE's to look at other ways of determining 'site quality' and what I am suggesting is that some of the 'signs of quality' I put forward would be well worth implementing if anyone wants to continue doing well in the SE's - which is certainly my intention!

If quoting google to prove a point helps then try:

http://appft1.uspto.gov/netacgi/nph-Parser?Sect1=PTO2&Sect2=HITOFF&p=1&u=%2Fnetahtml%2FPTO%2Fsearch-bool.html&r=1&f=G&l=50&co1=AND&d=PG01&s1=20050071741&OS=20050071741&RS=20050071741

its a long hard read but shows Google is looking at other ways of scoring sites, including links of course but covering other areas such as #38 deals with the domain name

Hegs - you are well informed - but why argue against the fact that SE's need to continully improve and find other ways to score sites determine search quality?

I would love to hear how you think they could improve search results?

Link to comment
Share on other sites

>> it is impossible even to determine which are the help pages

Not hard - If in anchor text "help" or "faq" or "support" and in linked_page (instring "?") then help page = true

If google can sort billions of docs in fractions of seconds then this really is not hard - I even do similar things in house here

The whole point (and my job) is to think one step ahead of the SE's - or try my best too - you can be damn sure 'the plex' 9googles headquarters) are working hard to improve result quality continously - you only have to look back a few years to see the results of not doing this - remember AltaVista?

Once such a giant and now provides 1 - 1.5% of traffic if you are lucky

Link to comment
Share on other sites

>its a long hard read but shows Google is looking at other ways of scoring sites, including links of course but covering >other areas such as #38 deals with the domain name

Buntina, if you want to know the facts go and do a regression of these factors against the search results rankings if you like. It's not so hard to do, and then you would be arguing from a position of science rather than voodoo.

>Not hard - If in anchor text "help" or "faq" or "support" and in linked_page (instring "?") then help page = true

Your false positive rate will be huge ("help with your homework" "athletic support" "faq on bananas"). Your false negative rate will be 100% for non English pages.

There are better ways of doing this using linguistic analysis - but they are still not adequately accurate (IMHO) for finding and classifying something as diverse as help pages. Government bills, yes. The search experts at Google, Yahoo, MSN etc do not work in a vacuum, they are all part of the wider IR research community and the techniques for doing this sort of thing and its limitations are well known.

Link to comment
Share on other sites

Hegs, these are simply ideas of factors that may be in place, or are a possible to be in place, in SE's that may determine the quality of a site, and yes with that simple 'if' statement there would be huge errors and I believe Google or any other SE would have a far better answer than that!

You seem to miss the point that 'indications of site qualtiy' is not a list of facts, more a list of ideas that could contribute to 'improved quality' and may improve your sites performance - seems to work for me 

Link to comment
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now
 Share

×
×
  • Create New...