Teoma Aims To Kill Google

gwernol writes: "SFGate.com has an interesting article on the relaunch of Teoma's search engine. They are trying to topple Google as the leading search engine. If their technology delivers on its promise then it will at least be some real competition for Google which can only be a good thing."

Only a search engine

[sirsnork]

While it seems fast and fairly relevent in the few searches I did. It doesn't have nearly the features of Google. There is not toolbar that I could find, no groups or images searches. Google is most certainly more that just a basic search engine these days!

For newbies its ok

[geeknews]

The consensus of users on my site when we posted this article was that Teoma will be great for less than net-savvy users, for more intelligent searchers Google is light years ahead of the rest.

Beta indeed..

[BelDion]

It'll be a while before this Teoma thing can topple Google.

First of all, no cache. The cache in Google sort of sneaks up on you in its usefulness.. Whether it's because the website is down or because you're looking at an html version of a PDF or word document, you find that you're using the cache all the time.

More to the point though, how friggin slow is Teoma? I hope it's due to relative newness or something, because it's frightfully slow when running queries. Google flies, click search and the page comes back next to instantly (on a broadband connection anyhow), Teoma seems to be taking several seconds right now. I'd say Slashdot effect, considering where we are, but what kind of poorly designed search engine crumbles under the slashdot effect?

But will they throw crap at you?

[iturbide]

Let face it. Just how good a searchengine is technically is only part of the story. The other part is how much advertising, cookies, links to 'buy a book about whatever on amazon' and all that will they throw at you? You get the idea. This is imho what killed off altavista and loads of other search engines. If people get annoyed enough, are thrown into a portal, or just plain have to wait too long for all that crap to load, they just won't go there.

If they don't get that right, Google has little to fear.

Let's Talk About This Tomorrow

[Schlemphfer]

Why are we wasting time talking about this search engine now? It launches at 5:00 PM Pacific time Monday. At that point, we'll be able to make useful comparisons to Google.

Similar results == Zzzzz.

[ahaning]

So they give different results. Good. If every search engine gave the same results, there wouldn't be any need for more than one. For example: Yahoo! or Google for searches?

Besides, this is still just a beta. No use in discrediting it until it's out of beta.

Ugly cheap logo

[Megs]

Okay, fine, they're allegedly going to bring out the Google-killing version tomorrow (News for April Fools, Stuff that makes for a really good laugh).

The real question is, are they going to get rid of that lame, butt-ugly logo that just screams "cheap knockoff"?

Also, in my profoundly unscientific survey of two friends on AIM, neither of them were able to correctly recall the name Teoma. Just because it means something cool doesn't mean that it will actually be a cool name...

Meghan

Ten Minute Searching Score

[sam_handelman]

I have evaluated a hit as relevant if it contains information related to the question asked. General information about Greece, or about the nutrient value of artichokes (but not containing specific info as to their vitamin content), I did not count as relevant. Pretty subjective, of course.

Query (relevant hits of top 5)
Google Teoma
Religious Intolerance by the Greek Orthodox Church
5 2(1)
Nethack 3.4 Spoilers
5 0
Vitamin Content of Artichokes
4 0
Average Velocity of Asteroids
4 0
Who won the peloponnesian war?(2)
5 5
Samuel Handelman Columbia University(5)
2 0
Harry Noller University of California Santa Cruz
4 4
Edward Dratz University of Montana Bozeman
5 3
Dangers associated with mercury thermometers
2 0
Did Turing have any children?
0 0
okay
Autobiography of Alen Turing(3)
5 2
Isaac Asimov's Middle Name(4)
3 2

Anyway, my time is up. avg. 50 seconds to run and squint at each query.

Subjectively, to all of these querries, the #1 hit on google contained the answer to my question (the EXACT vitamin content of artichokes, the NAME of the side that won the war,) while Tacoma, even though the hits were relevant to the question, it was not clear if the information I sought was actually in the returned result; except for my former faculty advisor and his colleague, which Teoma found just fine.

(1) I'm counting the Scientology hit as relevant.
(2) Google corrected my spelling, which Tacoma did not. I'll accept that from a Beta.
(3) Turing didn't write one. It was a trick question. Any link to a review, specifically, of either any of three (that I found) biographies of Alan Turing I counted as a hit.
(4) I didn't get his middle name, but it turns out he wrote a story called "Middle Name" which swamped the results. Google found specific references to the story, whilest Teoma returned links to lists of Asimov's fiction, but I generously scored both as hits.
(5) when I put my name in quotes Tacoma University either a) cannot find any matches or b) doesn't understand what the quotes mean. I assume b since none of the hits it finds without quotes mention me.

Anyway, I'm satisfied in calling that statistical signifance (95% chance) that google is better.

Re:I care nothing for Scientology or firearms

[Anonymous Coward]

Google didn't remove xenu because it was "offensive", they removed it because they are required to by law: the DMCA, to be specific. If you want to complain, complain to the lawmakers who allowed DMCA to exist in the first place. And before you say links don't violate copyright, I agree, but the DMCA has already been used to remove links in the past -- remember the 2600/DeCSS case?

And if you want to talk about spinelessness, why hasn't the owner of xenu.net simply contested the claims that his site violates the Co$'s copyrights? Google has said several times that that is all he needs to do to get relisted. If he's too spineless to do it on his own, he could ask the EFF to help.

Re:I don't think so

[optikSmoke]

Plus Google sounds cooler than Teoma :)

Exactly! I've said it before [slashdot.org], and I'll say it again: with a name like Teoma (or that other one.... Visivimo or whatever), nothing is going to topple Google. The name Teoma is just another product of today's pattern of "Let's choose exotic, foreign-sounding words so people will think we're *kewl*, man!". Maybe these people should try: "Let's choose normal words so people remember our name." At least that way, people won't be asking "Do you remember the name of that thingy that tried to surpass Google?" by this time next year.

The thing teoma lacks

[Cam Wheeler]

My favourite feature on google is the "did you mean" listing if you seem to incorrectly spell a word. I don't know about the rest of you but it's a fairly regular occurence that I don't know how to spell something complicated and I'll end up with no search results. Google helps me find the results and assists me in knowing the correct spelling next time.

Re:My study of Google, AllTheWeb, Teoma, and WiseN

[Anonymous Coward]

Given that these search engines differ in how they work, comparing the number of hits for a given query is not even a remotely relevant metric for how good they are.

The only way to compare search engines correctly is to perform a comparative analysis of the results returned. For each query you need to analyze if the search engine omitted pages it should have known about, if the objective relevance of the page corresponds to its placement in the result set etc.

Also, if you are going to "test" search engines, make sure you figure out how things work. For instance searching for what some people might consider offensive without first turning off filtering for potentially offensinve content and then just coundting the URLs doesn't exactly make you look too bright.

Your "test" is akin to someone who has never driven a car comparing the performance of sports cars by looking at their instrument panels.

Ooh, this one has a speedometer that goes all the way up to 300mph. Surely this car is much better than the one that only goes to 200mph.

Please turn on brain before use.

Re:Where's Teoma's caching?

[rseuhs]

First, you don't need a database because you just have to fetch pages, the search-index is either unrelated to this or needs it anyway.

Secondly, downloading gigabytes of data is not free, it costs bandwidth. Consumer-prices around here are about 0.05 $ per Megabyte, let's assume that Teoma pays 0.01 $ per Megabyte (Yes, I know that they probably don't pay on a per-megabyte basis, nevertheless they have to pay for their bandwidth one way or the other. If anybody knows how much this costs more exactly, please feel free to correct me).

To download 10 Terabytes would cost 100000 $, cheap IDE-harddrives cost about 2$/GB, so storing 10 Terabytes would cost about 20000, or 5 times less. (Please note that 2$/GB are retail prices, if you actually buy 10Terabytes of harddisks, I guess you will get some kind of discount ;-)

If you also take into account that you have to reindex sites frequently, (let's assume monthly), the yearly cost of operating the search engine is 60 times the cost of "storing the web".

So unless I'm completely off-scale with my assumtions, the cost to maintain a cache is actually neglegtible compared to the cost of basic search-engine operation.