fantomaster

msg:225634 | 1:06 pm on May 18, 2000 (gmt 0) |
Know what you mean: can look very spammy, even if unintentional, and may give the competition some nasty ideas. One way to work it is to set up Apache's mod_rewrite to redirect Alltheweb spiders to a 404 page and resubmit the pages you want to see dumped to Alltheweb only. May take a couple of weeks or so but should do the trick nicely.
|
engine

msg:225635 | 10:46 am on May 19, 2000 (gmt 0) |
Thanks Fantomaster - good idea - I can then selectively get the best (most relevant) pages listed. Cheers
|
Brett_Tabke

msg:225636 | 5:51 pm on May 19, 2000 (gmt 0) |
I had the same problem a few months ago, but I thought they'd worked this out with the algo itself. Many of my near identicle pages aren't showing up on the same page anymore. I thought they had it fixed...
|
engine

msg:225637 | 8:03 am on May 20, 2000 (gmt 0) |
Brett, no they haven't. It's embarrassing to have too many top positions. Fantomasters description is right, it looks very spammy. I suppose the pages are proving to be "more relevant" than any others, but one page showing up would be fine.
|
joanr

msg:225638 | 3:17 am on May 23, 2000 (gmt 0) |
Couldn't you simply use the meta "noindex" tag or does fast ignore it? I had this problem also with sub pages and the default.htm taking up five or more top 10 positions unintentionally. It did look like spam. I had to change titles/keys/descriptions on a few pages and some I just specified noindex. Funny how it's surprisingly easy to have this occur. One problem is using similiar descriptions, like a company byline, on several pages of similiar content. A newbie mistake and one which I paid for dearly on AV. I look at some of my competitors' awsomely designed sites, knowing that they paid a mint to have them designed, and sometimes feel bad that I outrank them, especially on Inktomi portals, but that's the way it goes I guess. Can't complain or the spider gods might hear me.....
|
VAL@Amsterdam

msg:225639 | 10:00 pm on May 23, 2000 (gmt 0) |
The same problem occurs when I am targeting many different keywords for a client. If I use a template / page for more then one keyword, chances are big to be listed more then once in the top ten-10 for ONE keyword, while the other keywords aren't ranking high at all. That was reason for me to optimize pages for ONE keyword (not more...), which makes it a lot easier to know exactly how many pages get in for a particular keyword.
|
fantomaster

msg:225640 | 12:26 am on May 24, 2000 (gmt 0) |
joanr: FAST/Alltheweb generally ignores meta tags, so your suggestion will probably not work. It's still ok for AV and most others (excepting Lycos), but not for Fast as far as I'm aware.
|
Brett_Tabke

msg:225641 | 9:06 am on Jun 26, 2000 (gmt 0) |
Well, a bunch of nearly identicle directory pages we had, have finally disappeared from the same page in Fast. The original is still there. It was really annoying. Are you folks seeing the same with yours? I'm wondering if Fast is using a low tech pseudo clustering algo now where they just check this result and only take the first from a domain. (admittidly, I've not checked real deep).
|
engine

msg:225642 | 4:31 pm on Jun 26, 2000 (gmt 0) |
Oh dear, still holding 1 through 6, 9,10,11, 13, 15 and 27. I'll keep an eye on it.
|
Brett_Tabke

msg:225643 | 4:29 pm on Jun 27, 2000 (gmt 0) |
I sure can't find anymore dupes of ours or in our kw categories.
|
seoboy

msg:225644 | 7:30 am on Jun 28, 2000 (gmt 0) |
engine, are the pages your getting dupes with subpages from the same domain, a-records, or c-names? seoboy
|
engine

msg:225645 | 8:13 am on Jun 28, 2000 (gmt 0) |
seoboy, hi, they are not dupes, they are genuine pages from one site. No special;) techniques are used to do this - it just happened. Perhaps the "themes" are too strong for the engine.
|
Brett_Tabke

msg:225646 | 10:51 am on Jun 29, 2000 (gmt 0) |
I don't think Fast is using theme indexing in any way that I can deduce. It is a relatively easy old fashioned algo. I'm telling you - it feels like the Infoseek 97 algo. It just occured to me what that would sound like to a newbie. The rich suculent spidering of Inktomi 98, with a full bodied indexer from Excite 96, and seasoned to taste with search spices from 97 Infoseek. lol It is a select crowd, this must be the place.
|
pete

msg:225647 | 12:57 pm on Jun 29, 2000 (gmt 0) |
Brett, That wonderful blend was before my time. I hear you guys talk about InfoSeek and when one could optimise, submit (and check ones ranking within a minute) and then re-optimise. Cannot imagine what that must have been like. Regards Fast: There was a big clean up on all my listings. Had a number of pages which were very similar in content and shape taking up all the results. Sort of felt a bit like a rudimentary duplicate elimination filter which cleaned it up. Regarding the timing of it, approximately 2 mths ago.
|
engine

msg:225648 | 5:59 pm on Jun 29, 2000 (gmt 0) |
Brett, i remember those days. Infoseek was "fun." Are there any angines out there offering "instant" listing? I've seen some offering instant spidering, but listing takes a little longer.
|
Brett_Tabke

msg:225649 | 1:44 pm on Jun 30, 2000 (gmt 0) |
None that I know of at this time. Gnutella will be interesting once it gets going.
|
seoboy

msg:225650 | 5:39 pm on Jun 30, 2000 (gmt 0) |
engine, w/in 3 days (ink) is pretty fast, IMO . sure beats the heck out of monthly. brett, re: gnutella - agreed. i like the way you think. :) seoboy
|
|