What's your favourite search engine that's not main stream?

pick an engine - any engine - from this list

[jeremy goodrich]

Which one of these would you say is a contender? Why / why not?

What characteristics make a good up and coming SE? What should we be watching for?

Some possibilities include:

Gigablast [gigablast.com] - Pretty interesting for a 1 person show.

WiseNut [wisenut.com] - Did they ever update their database after getting out of beta?

Vivisimo [vivisimo.com] - Interesting concept, too cluttered for my taste, but perhaps as you get used to it, it gets better?

Yuntis [yuntis.ecsl.cs.sunysb.edu] - Very interesting, loads of bells and whistles, university project is definitely one to watch.

Alexa [alexa.com] - Owned by Amazon, can you say big brother? (sorry). But - they got news & reviews on every site of the web...well, at least - they say they do :)

KarTOO [kartoo.com] - Neat flash interface, company says it's a product demo - kind of like Alltheweb is a product demo for Fast Search & Transfer. Another novel concept, but will it fly with the public?

Search Hippo [searchhippo.com] - Another 1 person show, they've got some interesting ( free ) keyword data, but nobody seems to be using it...will that change? It's been around for a while.

<added>
OpenFind [openfind.com] Forgot this one first time around, very large database, but the algo needs some refinement - as well as a refresh or two. In the same league as WiseNut, imho.

[edited by: jeremy_goodrich at 10:51 pm (utc) on Feb. 19, 2003]

[brotherhood of LAN]

While that is not always bad, how do we know the difference between a 'quality' engine and a small one that somebody built over the weekend?

Nice summary ;) You also highlighted the point of the maker of gigablast being a one man show, I guess that illustrates the one in a million chance that an "out of the norm" engine can shine.......it seems that "team of phd's" rings a bell when it comes to making a respected engine! No doubt if the maker of gigablast had another 10 of himself they could almost be a google

[papamaku]

I do agree with what you're saying and if this forum is intented for the 'almost there' sites like gigablast, kartoo etc

why not have a 'search engine spotting' style forum, for reporting new sites or features that poeple have come across

just to keep the 'always looking for the next big thing' people happy :)

[jeremy goodrich]

WiseNut had what, 17 when they were bought for stock by LookSmart? I know it was less than 30.

I'm not saying 1 guy can't build an SE...but, notice I also said the database has to be 'large'...:)

Consider also that the builder of Gigablast -> was a software engineer at Infoseek...has a bit of experience / history in the industry.

How many of us here have built an SE? I have built / modified my own :) as well as helped build another.

The database got to about 20 million urls or so...and the algo was getting good, but then I realized -> why would people use *my* search engine, instead of Google?

Nail that, and you KNOW we will be talking about it here. :)

For Gigablast, eg, it is the real time spidering with a db of 150 or more million urls...that's huge.

Hope that helps. :)

[Brad]

Jeremy I see your point, but it also reinforces the status quo.

Over time, there will be less and less to talk about.

[jeremy goodrich]

Only if innovation ceases, Brad.

And that, I do NOT see happening. Ever. :

[cohwill]

Jeremy,

I'd love to build a specialized search engine for a specific topic. And I even have the topic in mind, and another like it does not yet exist online. But, alas, am not a programmer.

NorthernLight still exists as a corporate intranet search engine.

Some of my favorites include, ExactSeek (an up and coming SE), DataCreek, IntelSeek, and WebWorldIndex.

[cjtripnewton]

Hyperbee looks really neat to me. I can't wait for Yuntis to kick into high gear.

[WindSun]

Hyperbee looks interesting at first, but the "copyright 2001" did not inspire much confidence. Nor did the fact that they say they will be testing searches from users "early next year", when "early next year" was spring of 2002.

[rbester]

DeepIndex [en.deepindex.com...] looks promising.

It is a rather small index but it deep crawls entire sites a few days after submission, and its crawler comes back for visits every now and then.

[cohwill]

Yes, but again DeepIndex states that its new search engine was expected to launch in December, but there is no sign of it.

[johnraphone]

I like Gigablast the best at of all those mentioned above. Gigablast is good with relavancy (could use a few fixes) and it looks promising, looks like a serious Google competitor in the future.

[karakas]

scirus [scirus.com], for scientific information.

[jeremy goodrich]

Specialized search engine powered by FAST -> not sure scirus fits in with the rest of the engines in this group. :) Too topical, too might have a future.

The rest of these, it's more like, innovate or die and I would love to see what manner of innovation they come up with.

Choosing a topic ( like Scirus has done ) and then running with it may just be the thing...if you can come to dominate that niche, it becomes a necessary tool for professional in that keyword sector.

Gigablast, OpenFind, etc all have their own algo's...Scirus algo is now OVER ( lol, since FAST was just bought by Overture.)

[mrdch]

Due to the TOS, I cannot discuss it - but I think it is relevant in this context. IMHO, Jeremy, I think we 'nailed it' :)

MC

Please read the TOS re: no self promotion, thanks

[edited by: jeremy_goodrich at 12:45 am (utc) on Feb. 26, 2003]
[edit reason] please read the TOS [webmasterworld.com] [/edit]

[DeepIndex]

Hi,

i just came by and saw this nice thread... and you were talking about .... DeepIndex. ;-)

Yes i do apologize, i stated starting end of december 2002 and that i would call back when runing... Sorry, i forgot to pay my phonebill, so i didn't call back on the day... Anyway, it's not my phone bill which is interesting us here ;-)

In fact the engine covers europe (partly) ;-) and also the algo is not perfect yet :-( . Rising up is not easy, as you stated here in webmasterworld... but when i launched the project last year i knew from the begining that this would be hard job, but i can insure you, it is a very exciting job.

So what is going? what is coming?

The natural crawling is still going on and also the free inclusion is working, but i prefer the paid inclusion one time fee (got to pay my phone bills and the servers). Meanwhile i prepare also localized searches (actually testing in france), as i believe that localized searches will come up some times.

Even if international trafic is actually not significant, the project is build to rise with time and i frequently talk with webmasters (also from here) to check out what should be done to make it more attractif to the pro users and normal users.

Best regards

Gilbert

[Romeo]

I have looked at the alternate search engines mentioned in the 1st posting of this thread.
Thanks for the pointers.
They are nice. Very nice. But ... my site never got much traffic from these, according to the following statistics for traffic coming from SEs:
84% Google, 14% Yahoo, 1% Altavista, and all the others are in the last remaining 1% (with Ask 0.5%, Mamma 0.1%).
This is sad, and I am not sure, if the arising of some sort of monopole is very good for the internet community in the long run ...

Regards,
R.

"In the long run, we are all dead ..." (Keynes)

[DeepIndex]

Hi Romeo,

even if their trafic might be low... these "might" become challengers...

They probably just need a little kick off...

Don't you beleave so?

Best regards

[GrendelKhan TSU]

wow, never seen KarToo before...pretty interesting.

Reminds me of www.thebrain.com though (visual mapping search). Wonder if they know about eachother.

[Brad]

While not a web search engine, one that I think is worth mentioning is Daypop. It seems to have quite a following amongst news junkies like me.

[manch]

I think AnswerBus (http://www.answerbus.com/) should be added to the list of best Alternative Search Engines. I doubt if Ask Jeeves is still on the Web next year.

[raven]

I've been reading over this thread and discovered some search engines I've never seen or heard of before; some which are pretty cool.

[edited by: jeremy_goodrich at 2:44 am (utc) on Mar. 12, 2003]
[edit reason] See TOS [webmasterworld.com] #16 Thanks! [/edit]

[IanTurner]

AnswerBus is a Meta, meta engines have never really taken off - the majors can crush them anytime they like by blocking feeds.

AJ on the other hand has Teoma, which is as good an engine as any at the moment, it just needs a larger index.

[KartOO4U]

I just want to add that KartOO is also a Meta.

[cohwill]

On the other hand, most of the metas are still hanging in there. Take a look at GreenManga.com. One of the engines it crawls is NBCi, and I don't know any other engine that looks at them.

[Fischerlaender]

AJ on the other hand has Teoma, which is as good an engine as any at the moment, it just needs a larger index.

This simple sentence includes one of the biggest problems for any search engine: Scalability.

It's easy to build an engine with an index of just a few million pages, but to expand this engine to hundreds of million pages could just be impossible - for technical as for spam reasons.

As for Teoma there could be three reasons for not having a lager index:
1. They just keep the index small to test the system out.
2. They have serious scalability problems.
3. They just don't care.

In other words: You cannot say engine A is as good as engine B, only its index is to small. This would include that it's just a matter of time or money to expand the index and this is not correct.

This means that we always have to look at the scalability of a newcomer if we want to judge how big and mighty it may become in the future.

[EliteWeb]

I enjoyed ViscaOnline until they ended their searchness. So now my next choice out of the list is Wisenut, however teoma is a good one also. Then i'll go with gigablast. :)