Search Engines of the Future

[mixjourneyman]

Omar: two words to solve your software/tech problems: gnu linux.

[Steve James]

Imo, it seems as if the perfect search engine would essentially know what one is searching for and tailor the results to that. To do that, however, it will still be necessary to comb through "all" the information just to determine which is relevant to the query. In itself, that is a massive task, which is essentially why search engines are successful. Put it like this, suppose there were no search engines, how would one use the internet? In the "old" days, before search engines, there was simply a green and white page with a list of sites. Everywhere you could go was listed, because every place was a specific server (often at a college). The other thing you could do was e-mail ... someone at another college, since no one had a computer at home or even a dial-up connection.

Of course, in a way, it was great; but it was limited. Yahoo (and some others) changed that by giving any user the chance to access (almost) any server/site with or without knowing what's on it or what he or she was looking for. Yeah, the first to take notice and advantage of that were the porn peddlers. Dmitri's right, imo, that Google does an impressive job nowadays of eliminating porn ads. (Then again, I don't go to porn sites ... er, anymore.) It's such an enormous amount of work that it's hard to argue with the need for some sort of income. Maybe that's the solution: i.e., have a "pay per search" provider or service.

[Darthwing Teorist]

That WolframAlpha thing is clearly biased. I mean, a search for Stephen Hawking returns results with a picture, and a search for Paris Hilton doesn't?! I mean, WTF?!

It shows that it must have an algorithm for good taste.

[Dmitri]

It shows that it must have an algorithm for good taste.

That'll be the day...

[Darth Rock&Roll]

A clever search engine only needs an algorithm that sorts through the appropriate responses to human readable questions.

For example, If I type in what is 5 x 3 google will return teh answer and show the work.

If i type in translate, I am sent to tranlator software specifically.

the bots help by moving in depth into the content of the sites.
if no htaccess block, the more content the bots get and the more new algorithms can work on it.

search engines never used to work on human questions like that and it is indeed moving closer to star trek everyday!

[bailewen]

Omar: two words to solve your software/tech problems: gnu linux.

I dual boot with Windows 7 and Ubuntu 10.4. I do not see how a different OS is supposed to affect search results. . . . oh, you mean the driver thing.

With Linux you get other sorts of things. I find that usually I just need to add the word "unbuntu" to the beggining of my string and then I get more relevant results. With Linux it is always about getting info on a piece of software that has a name that is overly generic like "mailman" or whatever.

I am impressed that google is starting to answer a certain amount of human questions. Still though, it's been a process of me getting more search smart rather than the software adapting to me. I've learned to use quotation marks for certain phrases or to add terms like "howto:" You can also add negative filters with minus signs. My favorite trick is when I see a relevant result I will often copy part of its content for my next search string. With programming questions this helps a lot. Understanding questions better would definitely be a big improvement. Sometimes my computer has an odd bug and all I can say is "(fill in the blank) is not working" and work from there.

[mixjourneyman]

what I meant was that if you ever want to solve a problem in Linux/Ubuntu/etc... all you have to do is search for your problem, look in the appropriate forum, where your question has most likely been answered and bam!
Also, for software, you can't beat synaptic package manager and it is all there, all free, and usually works as well or better than windows.
Life after windows has been wonderful for me

[bailewen]

Linux still lacks an adequate photoshop replacement and the video editing software is still a far cry from even Windows Movie Maker. So far I have not gotten photoshop to work under WINE so even though I generally prefer linux, I always have to go back to Windows on occasion. Further problems are that the subtitles do not display properly on my Simpsons DVD's on any of the Linux players, not VLC, not M-Player, nothing. It's ok for me but I like to use them with English subtitles turned on as classroom aids at school. That means I need to go back to windows for that. And then there's games. Again, I have read about playing Quake or Doom on Linux but never gotten either of them to work. Lastly, iTunes. Amarok is a piece of crap and I really don't understand why people like it. Maybe it's because I work in Gnome but it always crashes and never seems to successfully scan my entire music library. Rhythmnbox is ugly but functional. It's a shame it doesn't manage podcasting with any elegance or efficiency. That's why I use GPodder. I've recently switched over to Banshee for managing music but really, none of the Linux media things measure up to iTunes which, again, I have installed successfully in WINE in the past but no luck with my latest Linux box. It installs but then won't launch.

So yeah, Linux is da'bomb but it still has deal breakers for me. I tend to mostly stay on it but still can't afford to delete Windows from my hard drive. Not yet.

[dr.k.conor]

The 'bugger' of search engines is often as follows:
Part-1: the programmed engine itself, which can be original programmed script or clustering: collecting other SE results. Sometimes this can also be subject and language domain limited.
Part-2: the data, information resource; this has an effect on the SE methodology; best in semantics, best in graphics, best in scholoarly resources. Wikipedia is both as useful reference and a misleading 'blog'.
Part-0: not usually considered, but refered to here:
> a return of a 'semantic-string' that is NOT within the desired subject-topic's domain; superman /=super+man
> return of the 'usual' definition: aging is considered my most as a 'biological' process, but is does include chemical changes, both desireable and non-desireable: 'aging' of grapes to wine, 'aging' of the wine to vinegar. If the internet existed for young Albert Einstein, he would have very little relevant data/information returned for quantum relativity; he was ahead of his time, meaning the data and data-links did not exist to give a return on the query. The counterpart to this is that what you want to research is 'not acceptable' such as TaiChi- and Lao-tse and Chang Sanfeng are unrelated.
> return of nonsense or return of misleading sources; such as the Paris Hilton archives for the Hilton Hotel, etc. Thus
Google, even advanced Google, may be able to shave down the returned information, but still suffers from the weakness of the sources searched. Such as theory of everything may through semantics or through deliberate 'bombing' of info providers return everything you did not want to know of christianity and devine creative forces.
Part-00: Mentioned above is the CSE, or custom search engine creation, supported by Google, with this you can delimit or stack the 'prefered' references..you would want someone to see. The weakness is that you will be out of date before you finish and that the site-links you use may be broken or obsolete. Also, the links may have been absorbed into a 'new' owners search tools and corrupted or that the new owner wants $$ to view them. Thus, speed of change and stability of the data comes into consideration.
>>Personally, I use a customizable Browser with either: Exalead, AllTheWeb, GigaBlast and create my own definitions and glossary.