Part of the Problem on the web

[MikeGale]

Here's a rare find. A nice factual exercise to work out what is swamping the web with useless content.

It from Danny Sullivan who has been going for years with his Search Engine analysis ventures. http://j.mp/aoSbrk

In this case it's direct evidence of how Google popularity is polluting the web. Makes interesting reading.

I imagine we could all throw up a few articles like this if we took the time and effort.

[Albert Wiersch]

Hi Mike,

Very interesting info about chocomize indeed (had to use the word there... let's see if Google picks it up. ).

It is really sad that there is so much garbage out there. I often run into it when doing searches. I would think Google would be able to filter the non-sense out, out least a decent chunk of it, but I frequently run into junk that is gibberish or has been "scraped".

And I wonder how Google can manage to store and index so much data when so much of it is replicated/automated junk!

[MikeGale]

I sympathise a tiny bit with Google management. Their revenue comes from sites like these horrors, so they are between a rock and ...

In the early web, pages were often ugly but content top notch.

Now a lot of "average" people have come in with easy publishing tools. Result unsurprisingly is bucket loads of rubbish. I often know that there's good content on a subject but it's vanished behind listings of utter rubbish. Effectively unfindable.

The Google computations are, in my opinion, an extremely good way to do maths (as far as I understand them). Trouble is they're based on input that, for me at least, is garbage. So we get GIGO (garbage in garbage out). A little quality human feedback would be a lot better than measuring how many people got fooled by a great summary. (but we saw what happened to dmoz, horrible)

Then I've sometimes found a tweet at the end of a search. It's really hard to do anything useful in 140 characters, and I don't think one Google found tweet has ever passed the test.

Such is life. The web becoming more meaningless. Until the next revolution. A luta continua.

<rant-ends/ >

[Albert Wiersch]

So we get GIGO (garbage in garbage out). A little quality human feedback would be a lot better than measuring how many people got fooled by a great summary. (but we saw what happened to dmoz, horrible)

I've gotta ask - what happened to dmoz? I tried searching but not sure I found what you're referring to... and of course, I got a lot of gibberish in the results.

Anyway, I think human feedback could help with this issue a lot... but abuse of it would have to be detected.

[MikeGale]

Sorry about that.

dmoz is what I call the Open Directory Project. It's the domain dmoz.org that it operates under.

You probabaly know all this, but I'll give a quick pen sketch for those who might not.

The idea is a federation of human editors who rate and write reviews for web sites. It has echoes of Wikipedia but is much older.

It looked like a great shining hope when it first happened, but seems to have (deservedly) vanished from most people's radar. When you put "noodp" in a meta tag that pretty much says what people think about it.

My personal observations of it's output, as far as I can remember something that old.

1) A lot of the editors took on an area where they had a competing site / product. It makes some sense, they know the terrain, but you could see the bias time and again as you looked at what was written.
2) Even with those interested parties, many, probably most, of the reviews / descriptions were written by people who seemed not to have a clue. The guys are working for free and are overwhelmed so it's understandable.
3) If a site changed there was a way to get it reviewed again. My recollection is that the system was an utter mess, unless you like waiting months or years for a change. (I have a dim recollection of a paid route, which attracted a lot of flack.)

It's a good idea that these guys had a shot at. (Stumble Upon, Alexa, Digg et. al have also had a shot at it.) One day somebody might get it right enough!

[Albert Wiersch]

Thanks Mike.

I don't think I ever used directory sites much, except in the "very early days" when the Yahoo directory was big. After that it was all search engines. I still have recollections of lycos and hotbot. Now it's pretty much all google and a little bing now and then when I want something different.

I can see how a very good directory site might be useful now with all the garbage links the search engines bring up nowadays... but then again, people might want the latest news, links, and reviews and I don't see how a directory site can keep up with that and everything else that is constantly changing.

[MikeGale]

With the tidal wave of rubbish floating to the top of search algorithms the approach that gets genuine informed feedback will take over if/when it happens.

I'd be happy with 6 finds instead of 30 000, and identified (where it counts) by people whose opinion gels with mine.