The next generation search engine

[cubytes]

Hi,
I have created a new concept for a search engine and I need some help developing it. I know there have been a lot of new search engines popping up, but none of them have got it right.........yet!!!

This concept will come with an almost guaranteed exit strategy and it will be relatively simple to develop, because it is not some new fancy algorithm in fact it is nothing more then a very complex mash-up with a content awareness twist.

A simple one two three and your results are consolidated substantially.

So what is this concept?

well
I wanted to call it Semantix, but that name is already taken, so now I am going to call it "SEEK".

How will it work?

This concept has two sides to it:
results
and
bundles

The results side will be nothing more then an in depth traditional search with a little "vertical search" search engine feature thrown in for a good measure.

this will have the one two three mentality
one for the search term
two for the keywords and tags to consolidate the results and
three is where the users choose which content to include in their results ( thumbs, pics, vids, blogs, and more)

The bundle side is the revolutionary part of this next generation search engine that will change the web as we know it. In fact content and information will come to you in your results instead of simply linking to it. I have come up with several bundle concepts, but the cool thing about the bundles is the fact that users will be able to customize and make their own. A few of the bundles I have come up with are:

thumb bundle- search results displayed in thumbs that can be manipulated
newspaper bundle- a well rounded mix of content and information
magazine bundle- more content with less info
book bundle- all about the info
and more

What do you think?

[lisati]

I think the name "Seek" is already taken..... www.seek.co.nz & www.seek.com.au

[cubytes]

that sux

how about "see king"?

[Lux Perpetua]

It sounds interesting, but I'm not sure what your bundle descriptions actually mean. An example of a search result with the different bundle types applied to it might help.

In any case, I wouldn't be much help to you developing this. Good luck, though.

[nanotube]

interesting idea... but if you actually present content from other sites to your users on your sites, you may be running afoul of copyright issues (e.g. see what happened to google over its providing of news snippets on google news...)

so.. explain more what you mean with the "bundles" thing...

[cubytes]

The search will work kind of like this:

-the user enters a search term for instance "computers"

the user then enters key words like "dell, pc, mac, etc"

then the user choose which content to include in their results

for instance

pictures,videos, reviews, blogs, forums, web-site thumbs, RSS feeds etc

and then the results are fetched and displayed

the bundles are merely different views for those results

I imagine the newspaper bundle will kind of look like what a real newspaper would look like but it would be filled with all of the above results.

same for the magazine and the book

hey here is a new bundle
how about a link book that would resemble a phone book

what do you think?

[cubytes]

yeah I had a feeling copyright issues would be a drag but what if the results were compiled of only user submitted content and information

[steveneddy]

I'd like to read more about this.

[cubytes]

yeah well the next generation of web services will work kind of like what I propose as far as on demand content and information is concearned without having to actually go to that site for this information, basically a more rich search based RSS off breed but the question is how do you achieve this sort of proposal?
can you go through API's and license agreements or will I basically go bankrupt by consistently getting sued?

the majority of data on the internet is user submitted anyways

[cubytes]

nice somebody just hacked into my computer and it almost crashed I had to literally rip out the Ethernet cable
isn't that lovely? Windows is so pathetic!!!

were was I ?

on demand content and information in other words on demand material

that of course might be a little complicated due to legal issues

but my search engine will work

it is a one two three mentality

I have even designed the UI

I want to continue my education but I can no longer afford it, if you want to get rich quick this is the best idea I have that will make a rags to riches dream come true, but this isn't the only idea I have only the one that I think will be the quickest.

in the end my search engine is a front end to a complex mash-up, the results will be a lot richer but not enough to go against copyright laws
at least I dont think so, besides like I said we can always rely on user submitted content because we are the internet right?

[cubytes]

contact info

cubytes@gmail.com
digitaldumpster@hotmail.com

drop me a line

[CptPicard]

It's the implementation of your bundles that is going to be the hard part; from what I understand it either requires some heavy semantic web style markup in the sources or otherwise heavy AI to "understand" the content in order to mash it up intelligently according to the views you want.

Of course, a search engine requires a lot of infrastructure too.

Have you considered trying to get a job at Google? :)

[zabien1970]

You write you want to compete against google and yahoo, then someone hacks into your computer and almost crashes it? Hmmm, I wonder if you scared someone. Have you ever seen the movie Antitrust? Maybe you should start locking your doors. :)

[Wybiral]

Hmmm, I wonder if you scared someone. Have you ever seen the movie Antitrust? Maybe you should start locking your doors. :)

lol. Maybe you should upgrade your firewall!

[23meg]

The search will work kind of like this:

-the user enters a search term for instance "computers"

the user then enters key words like "dell, pc, mac, etc"

then the user choose which content to include in their results

for instance

pictures,videos, reviews, blogs, forums, web-site thumbs, RSS feeds etc

and then the results are fetched and displayed

the bundles are merely different views for those results

http://clusty.com

[cubytes]

check this out

http://www.readwriteweb.com/archives/search_20_vs_tr.php

but like I said no one has got it right yet

I heard of this intuitive search rumor that sounds pretty interesting basically a search that always runs in the background given you results that you may be looking for

the point is:

you don't have to make really significant changes to the fundamentals of search engines and there is no need to make them a social network either, but what I plan on doing first is basically a change of the interface layout,so I can snag some extra input from the user to consolidate their results to exactly what they are looking for with no need to have clusters or categories. I also want to make it look a little more visual appealing!!

the bundle idea was just the expansion of this concept to the next level which would be of course on demand content and information
but I have been thinking about how to bring that into fruition and I am onto something, but I just haven't got my mind totally wrapped around the idea yet

[cubytes]

Ok so here are three aspects that make the web come to you instead of you going www.blahblahblah


RSS
Stumbleupon
search engines

I had this idea once about distributing data, content, material and information all mashed-up together in a similar fashion as cable TV programing constantly streams to your cable box, to browse the mash-ups you simply change the channel.

Users will set up the mash-ups by choosing which content, information and material that they are interested in.

but I am still working on that one.................

the end result will possibly be a mash-up of RSS, Stumbleupon and search results, but then again that is just an idea

[CptPicard]

Push technology of this sort was attempted in the late 90s already by both Netscape and Microsoft, and they both bombed, in case you don't remember (or are too young to have been around) -- they even had "active desktop channels" :)

Perhaps it could be attempted anew with more intelligence about what the user wants to see, but don't forget that the Internet _is_ a medium where the user wants to be active, instead of a passive consumer of whatever is delivered...

[cubytes]

great point

I never thought of it like that

it makes sense

just a quick question

what was the name of those projects?

[pmasiar]

http://en.wikipedia.org/wiki/Push_technology

[cubytes]

thanks for the link

I see why those projects failed

thanks for letting me know a head of time..................................

[Midahed]

A bundle to exclude portals would be on my list....

Damn things just take up space in the search results and rarely add any value.

[cubytes]

cool I agree, what is the point of returning a bazzilion search results when chances are the first web page or two is the one your looking for, so instead I would rather fill the results with as much related content as I can, giving the user a more robust and richer set of results extremely consolidated and streamline search results based on the tags, keywords and descriptions the user gives in addition to the search term. Like I said it is a one two three mentality

1) search term
2)keywords, tags, and descriptions
3) selected content

and your results will be more personalized then ever before

here is a quick scenario

on the web page there will be the normal search field
the user enters the term "computers"
then hits search-
the search initializes in the bacground while a small vector window pops up right under the search field-
in this field the user selects what content to include in the search
i.e pics, vids, blogs, links, thumbs, etc-
the user hits enter and the search begins
when the search is finished another vector window pops up
in this field the user enters these terms "PC, GNU/Linux,freedom of software, etc; this field will be the tags, keywords and descriptions of the search term, the program will take this information and search for it in the search results-
then the user hits the go button, and the search is finished and the results will appear

[primski]

too much work.... average internet Joe tends to be lazy and doens't want to click all over the place just to get a search result.

It could be cool and usefull for a certain type of people (read, us geeks who spend like 90% time behind computer, not counting sleeping which is overrated anyways :P) but John Doe and GI Jane wants to get results as fast as possible and with little resistance as possible.

I like your idea, could be done great, but don't include too many options to choose form, by default it has to be simple.

Good luck mates

[cubytes]

you have a point

I will take that into consideration...............