who is brave enough to take on Google

[mshteam]

Or we could all put our knowledge together, to created the biggest index of underworld methods of money making and site optimisation.


...... oh wait.

 

[Mp3Mage]

Or we could all put our knowledge together, to created the biggest index of underworld methods of money making and site optimisation.


...... oh wait.

HAHA that's great. I just started some new methods from this forum and it's working wonders. I've made so much more money online with this forum each and every week I'm starting to laugh.

If you have enough money to build & create a quantum computer (yes, they exist already. google it.) and have the money and resources to hire the worlds greatest minds to provide a search algorithm to run on it, then MAYBE you might be able to compete with google.

most search problems are NP-complete. Solve this and you'll be given $1m anyway (literally, loadsa companys have offered this!), so yeah ...

Before I was studying switching my major from Comp Engineering to IT at BYU, I realized that there are so much to learn and it will take a lifetime to be able to understand alot of the things that probably google has implemented already. Great post Jammie

:usa:

 

[aмillionaírе]

i do this basically as a degree. AI is just pure searching and deals with issues like this.

Believe me, its no easy feat.

If you have enough money to build & create a quantum computer (yes, they exist already. google it.) and have the money and resources to hire the worlds greatest minds to provide a search algorithm to run on it, then MAYBE you might be able to compete with google.

most search problems are NP-complete. Solve this and you'll be given $1m anyway (literally, loadsa companys have offered this!), so yeah ...

get reading your favourite maths, physics and AI books, then in 5 years time you can have an idea of how complex google is, nevermind have your own search engine!

I think that's overly-exaggerated. It's really not so complex that you need to involve exhaustive physics in it. You only need an algorithm that would produce relevant products [the finish result, public product], and it only requires good mathematics to do that, also good programming skills to put the algorithm in motion [finish product]. Google is just a piece of long code with a good algebra algorithm.

 

[PhiltheBear]

Google is just a piece of long code with a good algebra algorithm.

If you seriously think that - can I interest you in a bottle of my famous snake oil?

BTW what exactly IS a good 'algebra' algorithm?

 

[fatjack'sblackhat]

Google started almost a decade ago with 2 guys and an investment of approximately 100 grand. I read that in USA today!!

 

[rjhere]

Taking Google on means creating a better interface and returning better results for joe user.. (not for you and I).. I really do believe Google's interface and results are about as perfect as it can get in the average user's eyes.. (our parents, non technical people, etc..)..

Ask.com tried to do this by fluffing up the search engine.. didn't work cause when it is said and done people just want a text based list of sites searched to from a screen with one field and one button.

Not saying it can't be done.. One thing that would be a google killer is a self learning search engine.. I believe Microsoft has this now but it's still in its infancy.. I've noticed that many of my techniques when new rank my high for a few months and then over time and gradually become less effective... That's a sign of a self learning engine.. Whereas google's approach is not gradual but an immediate change meaning an update.

Just my two cents..

 

[PuppetMaster]

Although nothing is impossible, this is pretty close to that. They have Ph.D's on their payroll and are way ahead of the game.

 

[r-webb-k]

I'm not trying to rain on anyone's parade, I'm just saying I'd be surprised if there was more than 5 people on this entire forum who are capable of understanding even the most basic math that goes into doing a simple query, let alone deterministic modeling, document quality analysis, or latent semantic indexing. This isn't something you can just pick-up from an ebook or a blog post. This is something that people go to school for years to understand the basic basic principles of.

cant believe you just said because people at BHW are some of the smartest people on the internet

 

[enjourni]

There are lots of search engines better then google... the problem is google has marketing power now. Just like the fact that Mac OS X is better then Windows XP, people still keep using XP bc of Microsoft's marketing clout. The "best" solution is sometimes the most popular, but not always. It's usually just the solution people are most familiar with that wins (it's true... people are essentially lazy and will often choose the familiar over the "better".)

Honestly for me I'm starting to be more interested in what the social bookmarking sites say are the popular sites for a keyword, based on the number of bookmarks. More and more I use delicious or digg to find new hot websites. Honestly I think I'm not alone... this is the future. Not that I particularly trust my friends to send me the best bookmarks, but that there are actual human beings rating the websites out there, rather then just whatever the googlebot thinks.

The web has gotten so entrenched with business and their ultimate elitism (how am I as a newbie website going to convince top websites to link to me unless I'm really good or I have money?) Social bookmarking bypasses this kind of crap because it lets real people do the actual linking (not just webmasters.) So you get sent straight to the site that is actually good, rather then just the site that paid money to have a link sent to them.

If someone is really able to capture what is going on in social bookmarking and put it together in a way that people can easily use, they might have a chance at google.

 

[jammie]

Nirvana, if it was easy, why would i be spending 7 years studying it?

Its too hard to design an algorithm that is effective in what it does.

If you have a spider that simply visits every page on the internet think of this ...

If there where 20 web pages interlinked, there are 2,432,902,008,176,640,000 different different ways to view those 20 websites. How do you decide which way to do? Server ping? Then some never get visited. Big no-no. Highest rating? then lower sites don't get visited as often. big no-no. Use probability? Use different heuristics?

Every search algorithm has problems. Its easy to check a search is correct, but checking if its the best is almost impossible.

Image 25 websites. Thats 15,511,210,043,330,985,984,000,000. Only 5 more sites interlinked and ALOT more problems already.

Imagine googles database of 60 BILLION webpages. try searching that efficiently.

Then you have sorting. Some searches run in Big-Oh Of n^2. So a search of 60 billion documents, would require 60billion^2 which is:

3,600,000,000,000,000,000,000 calculations (worst case)

thus, searching this way isn't effective. Then you have about how to rank results, what data to collect etc.

Trust me ... search algorithms SUCK!

 

[twinkle88]

I'm not trying to rain on anyone's parade, I'm just saying I'd be surprised if there was more than 5 people on this entire forum who are capable of understanding even the most basic math that goes into doing a simple query, let alone deterministic modeling, document quality analysis, or latent semantic indexing. This isn't something you can just pick-up from an ebook or a blog post. This is something that people go to school for years to understand the basic basic principles of.

Where can I get this book
Or
You said that the book isn't released yet, where did you get it?
Thanks.

 

[seomanifest]

Wow taking on google that is quite a statement, I wonder how many people here spend most of their time looking at other people's way of making a quick buck and instead of actually implementing these ideas they get lost on the forum reading and downloading the latest crack for the greatest bh tool which they probably never even use after they download it. I am not saying everyone here is like that, because obviously there are some very smart people here it's just that sometimes we miss the point of the information that is being generously offered here and get lost in our insatiating need for new knowledge that we never put into practice.

 

[jammie]

you can get any book off amazon.

Books you might be interested in:

hxxp://www.amazon.co.uk/Discrete-Mathematics-Applications-McGraw-Hill-International/dp/0071167560/ref=sr_1_3?ie=UTF8&s=books&qid=1224791578&sr=8-3
hxxp://www.amazon.co.uk/Artificial-Intelligence-Modern-Approach-International/dp/0130803022/ref=sr_1_1?ie=UTF8&s=books&qid=1224791620&sr=1-1
hxxp://www.amazon.co.uk/Introduction-Automata-Theory-Languages-Computation/dp/0321462254/ref=sr_1_1?ie=UTF8&s=books&qid=1224791677&sr=1-1
hxxp://www.amazon.co.uk/Distributed-Systems-Principles-Andrew-Tanenbaum/dp/0136135536/ref=sr_1_2?ie=UTF8&s=books&qid=1224791717&sr=1-2

There the basic's, i've read them all and know they will give you a pretty good understanding of the basics.

Once you've got through those, if you need anymore, just let me know and i'll give you the intermediate stuff!

 

[antx16]

now that is a mega bookworm , did you really read all that jammie , if so well done , did you learn anything from that that would help this topic ????

 

[mattstrike]

Going up against the big G is tough, because they hold so many of the cards.

Jammie is spot on when it comes to algorithms sucking, but if you take a look at what G is up to (ana1ytics for instance) you can start to get an idea of how they can combine algorithmic search with other factors. If there were two sites, for instance, that both scored about the same from the algorithm, but one had a much higher bounce rate than the other, which would you guess is more relevant to the people searching?

IMHO, the next generation of search engines will have to hybridize these kinds of approaches, balancing algorithms with indirect "voting", and adjust the results accordingly.

Could someone else try to get started? Sure, but come on, we all know that G has a huge head start because they already have a decent algorithm, toolbars in lots of browsers and ana1itics on half the sites out there.

 

[buttercup]

Well guys sorry i did not finish my thread , the power went down , anyway what i meant was....

How? How the HELL did you hit Submit Reply if the power went down? Very interesting ability, it could be useful, care to share? I mean, you certainly cannot be implying that you intended to edit... It's the first post lol.

anyways, sincerely looking for the magic mental post button. I would like to work while I ride bicycles and fly kites. Is it on my keyboard? The back of my PC? It's not inside, is it? I don't really feel like sorting through my messy wiring job ^^

 

[antx16]

Going up against the big G is tough, because they hold so many of the cards.

Jammie is spot on when it comes to algorithms sucking, but if you take a look at what G is up to (ana1ytics for instance) you can start to get an idea of how they can combine algorithmic search with other factors. If there were two sites, for instance, that both scored about the same from the algorithm, but one had a much higher bounce rate than the other, which would you guess is more relevant to the people searching?

IMHO, the next generation of search engines will have to hybridize these kinds of approaches, balancing algorithms with indirect "voting", and adjust the results accordingly.

Could someone else try to get started? Sure, but come on, we all know that G has a huge head start because they already have a decent algorithm, toolbars in lots of browsers and ana1itics on half the sites out there.

Maybe by looking into their new browser chrome that might give a few tips away, i am sure that browser is not just geared up to be a simple search browser , i am sure they have something else up their sleeve

 

[eduardo]

This thread made me chuckle. Microsoft has spent over 1 billion dollars trying to "beat" Google. So yeah, good luck guys!

 

[nundinator]

This thread made me chuckle. Microsoft has spent over 1 billion dollars trying to "beat" Google. So yeah, good luck guys!

Yahoo is about to lose 1,000+ employees. Maybe one of them could help.

 

[Morpheus]

Take on Google? Could be done. They didn't invent the search engine. They innovated in a way the world saw useful. Two smart guys and 100k got the ball rolling. They came out of "nowhere" and went up against king Yahoo!. Yes two guys and some good innovations started a company that outsmarted and dethroned billion dollar giants. Can it be done back to Google? Absolutely, however NOT EASILY. Larry Page and Sergey Brin (Google founders) are quite aware of what they did and how they did it. I'm sure they (and their tens of thousands of employees) have there "feelers" out there in every direction scouting for any considerable threat to their dominance. That being said, don't expect to ever take out Google. You're best shot would be to capture a good share of the market. Yahoo! is still around, and going fairly strong. Best thing to do is create a new "game" and be the only player.. An idea so innovative you could capture the market and be many many months ahead of the nearest possible competitors. Google is worth I don't even know how many hundred billion dollars now.. you'd better believe there are THOUSANDS of companies with FUNDING trying to come up with superior search technology to get their share of that market. Whoever holds that crown pretty much owns a "gold factory". I'd be willing to bet that it happens again. That out of nowhere there becomes a major player in that industry. I'm also willing to bet that it isn't any corporation with massive funding, but once again a couple of smart guys with brilliant innovation who were able to think waaaay outside of the current box the search engine industry dwells in.