Saturday, August 20, 2005

Social Search (un)Defined?

Or, using a C idiom, how about #undef Social_Search

When, I first started contemplating the idea of a social search engine, I only had a very vauge idea what it was, let alone how it goes about doing what it is supposed to. Even as I write this I have only a vauge idea of how I am going to do it, but atleast I am a bit more clear about what I really wanted to do. Infact what I want to do is very far away from what the term 'Social Search' may connote, and henceforth I shall relegiously eschew from using that term, lest someone think I am doing an online dating service of sorts.

So, what is that I want to do with this? Well I want to search blogs. What's new there you ask? Technorati does that, and why even the omnipotent mother of all search engines google does that, so whats new here. Well whats new here is that I simply don't want to stop with searching blogs, but I've realised there is something more that one could do with blogs, that you can't really do with the web as a whole (or its just too darn intractable).

The key here is that blogs are a deluge of semantic information. The previous sentence is so loaded with meaning that I've decided to allocate a seperate post about that. Coming back to what I want to do is that I would like to mine this semantic information, and try to present it in an web-based format. Not only should what I envision be able to do dumb old reverse indexed keyword searches (the quick search), but also must be able to use the abundant semantic information available on blogs to provide more meaningful results.

So, when do I want to this? Not immedietly, may be its a viable final year project, maybe its not. So, lets see how this comes about.

On a totally unrelated note, I tried to gork the details of 'The Java Virtual Machine' (blue-book?), and ended up disillusioned. I was never a big fan of java (atleast not much as ruby/python), but the JVM is clearly one of the least elegant (see, I am not stressing efficiency here, of which I have no idea about) designs I've come across. And from, what I read here, I guess the guys at Microsoft have got it right this time around. So, should I swallow my pride and give them a 'pat on the back' for the good work, or should I just silently ignore that ever happened and play spoil sport? Tough one...

PS : With 'Sriram Currying' in C# and all that, it seems more and more like Greenspunning, doesn't it?

Signing Off,
Vishnu Vyas

2 comments:

Sriram said...

I knew it would only be a matter of time before someone invokes Greenspun's law. Btw, that interview from Anders is from 2000 :-)

Do ping me when you see me online - I've been thinking of social search engines too - but from another perspective

FuzzyLogic. said...

I know, its a bit dated.. but well the point was to say they got it right from the beginning.. :).