Slashdot Mirror
Search Beyond Google
An anonymous reader writes: "'Search Beyond Google', the cover story of the March issue of Technology Review, is one of the few current Google stories that discusses whether their technology can stay ahead of the competition in the months to come."
They key for google providing relevancy is certainly eliminating "search engine spam". Almost everything that comes up on the first page for most things I search for is a referral program selling either something I'm looking for information about, or selling something completely different.
500GB of disk, 5TB of transfer, $5.95/mo
Google has had the last few years virtually unchallenged as the #1 search engine, because nobody has yet come out with anything that's better than PageRank.
But, five years is a long time to sit on an innovation without making it better. It gives the competition time to catch up. Furthermore, since PageRank doesn't seem to have seriously changed much, it's actually slipped backwards a bit as more and more people have figured out how to "beat the system" by posting nonsense sites with links to the site they want on top. Google's clearly trying to fight this, but that's an uphill battle.
Meanwhile, Yahoo now owns three distinct web-crawl based search engines, AltaVista, AllTheWeb, and Inktomi. They also own Overture, which begain life as GoTo.com who was the first to associate real search results with targetted ads. Put all these pieces together. Yahoo also has the original mega-directory site, which Google tries to duplicate by presenting the Open Directory Project on their site. In short, Yahoo's got all the resources to launch a brand with everything that Google has going for it... and when you look at AltaVista and AllTheWeb they feel quite a bit like Google already. Clearly, Yahoo's gearing up to issue a challenge to Google.
It really seems like Yahoo is making sure they have all the tech in place right now. When they're sure that they're better to Google, I fully expect to see a marketing campaign claiming that and inviting people to do head-to-head searches.
Google, as it stands now, is going to look pale in such showdowns. They've got to seriously modify PageRank so that the link spammers get downranked before Yahoo issues that challenge, or else Yahoo could reclaim the search market under it's "Google-killer" product line, and then direct people back to the original Yahoo site for their other portal needs.
Hopefully google will not go public anytime soon like they were talking about earlier. I fear that this would stifle their innovation and bring it closer to some of the other failed portals.. ie more ads in an attempt to satisfy investors.
I think it is a good idea for other search engines to step up to the plate and challenge google. It stops them from beoming complacent and spurs innovation from a desire to be #1.
I see no reason why the cycle cannot repeat. In fact, the cycle may be much like the semiconductor memory business, which has seen boom-bust cycles every few years since the early 70's. Sometimes a name will ride out for many cycles, but usually the company (and as necessary the technology) behind the name changes radically.
...maintain their technological lead, goodwill toward them will give them some breathing room. I continued to use Altavista for quite a long time after Google came out. It was what I was familiar with, I liked it, and it worked. Why switch? Eventually, I realized that Google had keen "read your mind" powers and finally switched. :-)
Javascript + Nintendo DSi = DSiCade
Every couple months it's "Can Google stay ahead of new competitor x?" And so far, everytime, the answer has been yes. People shift from search engines quickly when they no longer work, and people are still heading to Google.
I type something in and it spits an answer back at me.
As long as that answer is in the first page, usually the first three items listed, people simply will not care about the backend technology.
MS and others will brag about the vastness of the numbers of matching items they can find; most people only worry about finding one or two sites.
This is going to be a big non-event...mark my word.
It seems like Google is starting to admit that they've hit a wall at improving their search technology, so they're starting to expand into other portal areas to anchor themselves down the same way Yahoo did when their directory-search model hit the wall.
But Yahoo seems to be investing in several of the surviving web crawlers from the early days. Clearly, they see Google's hold on the title as the #1 search engine as something they might be able to take back.
Enough branding studies have shown that it's very very hard to knock someone off their post once they seize a certain mindshare - e.g. Coke, Windows(grin), and now Google.
So, irrespective of the technical competence, or otherwise of Google, it is going to be around and the leader, for a long time to come. P.S. My favorite missing google feature: search for bittorrent files
Google has been successful due to original thinking. It needs to ride its wave of reputation now rather than later in order to snatch up some of the finest minds to stay on top of this industry that is all about originality and fresh ideas. They seem to be on the right track by providing the work environment that they do.
But no more stuff like that Friendster wannabe site.
I've always been a google fan, but this article is essentially dated on its release, given the fact that the Yahoo! switch has already occured.
I do hope Google can continue its innovation, and reduce much of the annoyance of bad results through blogs.
I'm suprised more attention wasn't given to the Google IPO, and what affect that might have on the "relatively small" 1000 person company.
-m.
Once a really great tool goes "commercial" it's all downhill from there. One of the main reason I switched to Google back in the day wasn't because it was fast and accurate (which was great) but because it had such a clean interface. Now there are sponsered links that clutter things up. And who knows when/if popups will be a necessary evil to "stay in business".
Are you Corn Fed?
Exactly. When Google falls behind, you'll know it because you'll be using something else. This kind of "Entity XXXXXXX may suffer setback YYYYYYY any day now" story isn't reporting at all, it's speculation and ghost stories.
Google should leverage that freaking huge database that they have and dejanews. None of the other competitors have that comprehensive of a set of data. Don't screw with it by adding useless features.
--pete
.....in that everyone uses it, and everyone HAS used it for the past five years, or longer. People trust it, and that is something that just doesn't vanish. Plus, they HAVE done new things, such as google news.
Someone should invent a search engine with regular expression support. *sigh* A world with regexp-enabled search engines... That would be a wonderful world to live in.
GAAH! MY PRINTER IS ON FIRE!!! PUT IT OUT! PUT IT OUT!
Google's market position when they IPO has nothing to do with their technology. It has to do with their brand. "Googling" for something is the effective equivalent of going to get a Kleenex. Noone asked for a tissue. The market is going to be buying faith in the Google brand, and it's loyal userbase.
-- http://www.criticalassets.com
It would be the rebuttal to Google bombing... searchers could fight back by giving the crap a thumbs-down. Of course, then you would have the bombers voting down all the ligit sites. Dammit.
The Philosophy of Liberty | lewrockwell.com
The majority of users who use search engines are just end users anyways and appreciate the simplicity of Google's page design. I go to Yahoo, Altavista and Lycos and there's half a million links all over the place. I go to Google, and there's a nice clean page with the text box smack right in the middle.
Visual appeal still counts.
The good thing about Google Labs is that the innovations are done voluntarily, on the engineer's own time.
...
But yes, the tendency to expand into new (and unrelated) realms is annoying. When Google starts supporting e-mail, as jwz's law predicts it will, we'll know we're in trouble. Oh, wait
Xbox reviews.. We think they're funny.
Just because you have a PhD does not mean that innovation happens instantly. Research proceeds at its own pace, and you can only go so fast. There are very few problems out there that will actually buckle when you throw more talent at it.
I don't really care who has the most advanced search capabilities. I use google because all the paid links appear off to the side in a different color.
Thats all I really want . . . to get my search result seperate from the commercially paid for product placements.
--Tsiangkun
Perhaps they're trying to be the central information aggregator? Many of their initiatives, webpage search, news search, usenet search, store searches... have to do with sifting through more information than humans can possibly handle. The Blogger acquisition and the friendster thing could be seen as peripheral endeavours that may yield a profit, but also might yield information on how to sift for relevance. e.g. handle blog relevance by studying interpersonal relationships of sites like friendster. After all, their web search is based on relationships between webpages of a sort. In that vein, Google Answers could be interpreted as an experiment to leverage the power of people in finding relevant results.
:)
Or I could be reading too much into what is otherwise standard corporate behaviour.
Personally I put google above any of the other search engines out there, its clean no nonsense interface gives you what you want. I am not interested in having a search engine have about 700 billion different features that have nothing pertaning to what I am seeking, Yahoo may have some nice tie ins but I dont need a page full of ads to get in my way or an advertisement for YAHOO SMALL BUISNESS.
You have been sig'd
But then it is not SPAM anymore is it? SPAM is unsolicited commercial email. If you are aggregating it then it is no longer unsolicited.
Magic Eight Ball: Outlook not so good., Hmmm, how about Excel and Word?
One issue I see is the user experience context.
I'd say that 80-90% of the time when I want to do a search, I'm usually, at that point, doing something on the web. Having Google reachable at another website is convenient compared to having to fire up another client application to do a search.
I think one feature that would be awesome in a client-side app (but wouldnt' be limited so) would be a user history of Google searches. I'm looking random things up all day and it would be cool if I could have those searches logged and organized for later review in case I want to look back at the searches I've conducted.
"Injustice anywhere is a threat to justice everywhere." - Martin Luther King, Jr.
Also the risk-taking will drop off a cliff once they are public. The litmus test for new products is much more stringent once you have quarterly reports. And yes, they are going to have an IPO, stop debating it.
Yes, Google has a spam problem. It has been getting worse over the last year. In April, 2003 Google stopped crawling the web once per month, and then recalculating PageRank based on that monthly crawl. Since then, there has been a question of whether PageRank can even be calculated accurately by Google.
I speculated about a 4-byte docID overflow problem in an essay last June at Google Watch. In recent months Google started a "Supplemental Index" for some curious, unexplained reason. Their total number of pages indexed was recently updated to 4,285,199,774 -- just below the maximum for a 32-bit integer. It looks as suspicious now as it did last June.
Last November, Google began using an on-the-fly filter to further refine the search results for ecommerce sites. Some spam was deleted, a lot of other spam took its place, and a lot of mom and pop ecommerce sites were dropped inadvertently. Many people were unhappy.
Further evidence that Google's old ranking system is broken is the fact that three famous Googlebombs, "french military victories," "weapons of mass destruction" and "miserable failure" are all still working. The first one is eleven months old. It used to be that such Googlebombs were suppressed at the next monthly crawl, when PageRank was recalculated. Now it seems that suppressing them is beyond Google's ability. How else can you explain why Google puts up with these widely-publicized embarrassments?
Google's results remain unsurpassed for noncommercial sites from EDU, ORG, and GOV domains, however. Their crawling of the noncommercial sector is the most complete of any engine. The reason Google does so well here is probably because spam isn't much of a problem in this area.
So far Yahoo doesn't appear to be making much of an effort at covering the noncommercial web. It should be added that Google has more of a spam problem simply because spammers have been focused on Google for so long. Once Yahoo gets the same attention from spammers, then we'll be able to make a fair comparison of Yahoo with Google.
FWIW, in Mozilla Firebird, you can select a bunch of text, right-click on it, and go "Search the Web"... . I've never had to open a separate window for searching. Now, it would be so nice to have this in other apps.
Take email, for example. My idea is that when I'm posting a query to a mailing list, as I type in the words, the program should dynamically build a set of "related links" for the content I have typed in the email. That way, people won't have to ask me to STFW everytime I act clueless and send a simple query to the list.
Alright, I'm kidding. I'm not a clueless user, but you get the idea. For any content on my screen at any given time, I'd like to be able to access "related content" from... er... a sidebar on the screen?
Actually, no one can buy a privately owned firm against its owners' wishes.
Hostile takeovers only happen to public companies.
What about KartOO, which visually maps out relationships between sites? At the moment it's a meta search engine (the beauty's in the visuals, not the out-of-date results it gets from AllTheWeb and Lycos), but if it became the new way of looking at Google's results I think it'd be the Next Big Thang.
$ echo "ceci n'est pas une pipe" | sed -Ee 's/(eci n|pas )//g'