I am writing this post to clarify my thoughts on the social search space where my project stands among all the others out there, what social search means, and my philosophy on where it should go.
So to start where does social search stand? Well we have some engines, Mahalo, Wikia, etc (help me out add links the comments). We have some platforms, Yahoo SearchBoss, Google Custom Search Engines that well not social search engines are capable of being extended socially. Then we have plays by the major search engines, namely Google Search Wiki. Lets try and review these in turn.
The Sites
Mahalo
Mahalo is a social search engine in that it lets users of the site build "pages" around search results. Users of Mahalo can add links, "fast facts", categories and more. Mahalo recently introduced Mahalo Answers, a system by which users can ask questions and Mahalo users can respond (very similar to Yahoo Answers), there is a payment system in Mahalo dollars that can encourage the answering process.
Mahalo's focus seems to be around the "common query" space. If they can define good pages around common queries then they can handle the bulk of the search traffic and show human parsed, edited and validated links, that in theory should be better than any algorithm for finding good content related to a query. The Mahalo Answers system ties into this by letting users find out about specific things not covered by the common query space or not covered on a Mahalo results page.
Wikia
Wikia has much the same focus as Mahalo, building wiki's around common queries. Wikia does not currently have an answer system built into it. Everything else said above for Mahalo essentially applies.
The Platforms
Yahoo Search Boss
Yahoo search boss is not a social search site, it is an API into Yahoo's search system that allows social features to be built on top of it. OneRiot is a good example of this. OneRiot is a site where you can see the "popularity" of a given search result overlayed onto the result. The whole site is driven by the Yahoo Search Boss API.
To get to the most basic crux, the Yahoo Search Boss API allows 3rd parties to start appending meta-data to the Yahoo results, including social aspects.
Google Custom Search Engine
Google Custom Search Engine's are a different beast than the Yahoo Boss system. Google CSE's allow builders of the CSE to limit the web that is being searched over. While not social to start, this simple concept allows 3rd parties to build CSE's for specific users, essentially socially limiting their search. I had a project about 1 year ago that did this, called iPrecis.com, it turns out to be relatively hard to figure out how to limit a person's search, without a lot of setup by the user (in my opinion).
Big Player's Endeavors
Google Search Wiki
Google Search Wiki is a recently introduced "overlay" on the Google search results page. Basically it allows you to "move" results around for your own searches. If I did a search for "James Ostheimer" and my blog came in at the bottom of the page, I could push it to the top. I could also write a comment about the search result right on the search results page.
For now Google is not using this user defined info for much (like changing search results ordering). In the future they may integrate this information into their algorithms.
My Thoughts
I think that all of these products have their place. I like the crowd sourcing idea behind Mahalo and Wikia. Yahoo Search Boss is creating a lot of innovation with it's full API into Yahoo search. Google Custom Search Engine, well having a lot of potential, I think is turning out to be somewhat of a dud. It simply hasn't grabbed hold of being a great platform to build on, mostly, imho b/c it does not allow the reuse of search results in a full fledged manner as Yahoo Search Boss does. Google Search Wiki seems innocuous to me until Google decides to actually use the user data in a meaningful way.
This all being said, I think there is a huge hole here. All of these services put the macro "crowd" into use here, while ignoring the micro "crowd". There has been a rapid emergence of social networking sites that are part of the "live web". Facebook, Twitter, FriendFeed, Jaiku, Identica... all of these sites are about live or near time interaction with a users social network. The micro "crowd" to an individual user is his/her social networks as defined by whatever services he/she is on. These micro "crowds" can be ten to hundreds to thousands of people, and the key to this "crowd" is that they are trusted more than the macro "crowd" of anyone on the Internet.
An example will help illustrate this: Lets say you want to get a new mortgage for your house, you have questions about what type of loan to get (apropos I know). You ask a question on Mahalo about this and get a number of answers, some answers say get an ARM, some not. You ask your Facebook network the same question, you brother says "if you get an ARM I won't speak to you anymore". Which do you go with?
So yeah, the above is contrived, but another example, you want to know the best way to fix a memory leak in some Javascript code. Would you go with a stranger on Mahalo with no knowledge of you or your skill level? Or do you go with the guy you work with who you know is a genius?
This is the power of micro "crowds", TRUST. You trust your friends, co-workers, family, and even acquaintances more than you trust some random person on the Internet.
This is a hole re-searchr is trying to fill. It will try to tie into as many social networks as it can, and become your glue for getting answers to questions. Notice I said glue, re-searchr is not trying to supplant your social networks. To use a computer analogy, your social networks are the operating system, and re-search just wants to run on top of them.
Right now re-searchr is actually doing more than this, because it was originally conceived as a platform for re-analyzing search results. I am right now trying to figure out whether that blends into the micro "crowd" concept or whether I need to split out these products (comments please!).
