building a human powered search engine

[pk698]

I guess we can approach this discussion from 2 viewpoints:

1. what is wrong with the current search? (we have a lot of people "gaming" the system so they can get on the 1st SERP)
2. is there a better way to archive the web? (rather than relying on a non-human algorithm)

I've been interested in building a human powered search engine for a long time now. my first attempt was in 2004, and didn't turn out too well. but I've had several years to think about what I can add to it. I read this good article to get more details about my subject:
http://www.ariadne.ac.uk/issue54/search-engines/

So, I was wondering if anyone would like to offer some tips about how I should approach this. I know some of you will say, "it can't be done." But for those who are more optimistic, please chime in.

 

[Luke F]

it can't be done

(At least not on a reasonable scale)

 

[pk698]

(At least not on a reasonable scale)

I don't know... what is "reasonable"? who should say "ok, that's reasonable enough"? I think reasonable is a relative term. Another thing is "scale" - which is quite big if you sum up all the stuff on the internet. It would be extremely difficult, yet possible. Just look at the internet: was it built in one day? It took a lot of time and effort by countless people. It's quite big, and continues to grow at a phenomenal rate.

If you have a pessimistic and negative outlook on life, this is not an endeavor you want to tackle.

Anyhow, to make my approach easier, I can start small and narrow my focus on a certain topic. For example, I'll use this site: xenforo. OK, so now we got xenforo. Who knows what it is? what does it do? who does it serve? how much is it? how was it built? etc, etc... So now we got a starting point. A lot of people know this topic, and I'm sure they are willing to contribute to this xenforo list.

So, maybe I can start small - while having big goals.

 

[The Sandman]

I know the guy behind Splore. I don't think it will be a human powered search engine per se, but it will certainly involve lots of human interaction.