Today our team delivered a new version of Sociotoco Search. We’ve worked for months on this little baby, and we firmly believe it is now the best people search tool in the world. Of course, that’s just our opinion. But we have good reasons for this claim. Here’s why:
- For most social networks we use multiple engines. For example, the Find People feature on Twitter sucks like a baby’s thumb. (For many months it was unable to find me in their system. Go figure!) So we augment the crappy Twitter search results with various searches on Google, Bing, and Yahoo.
- We use real-time crawlers to trace links between profiles on different networks. For example, even if you have a ridiculous Twitter name like lemoentje, and you have a normal name on LinkedIn like Jeroen van Eck which includes a link to your Twitter account, then we will still find you on Twitter! (Even when your normal name is not mentioned on your Twitter page.)
- We are building this thing as a social learning tool. When people connect profiles (by adding them to the top bar and choosing some action like Open profiles) we will store that selection. This should improve the search results when similar names are searched again later. (Well, that’s the basic idea anyway. Doesn’t work yet.)
- And of course, our people search tool is the only one that looks as if someone had some experience with design. (But I readily admit there’s still a lot to do, which is the reason we’re calling it a pre-alpha release at this time.)
Since the first (silent) release last week we got some very encouraging feedback. The only real problem some people report is the fact that we created the GUI with Silverlight. Because Silverlight doesn’t work (well) on Linux or Mac.
So, why Silverlight?
Because our tool is free, and we wanted it to be fun to build. And (for us) neither Ajax nor Flash is fun.
If it’s free, why did you build it?
Because we need hands-on experience in working with all the major social networks. We now have experience with caching problems with Flickr, alternate layouts of MySpace screens, the LinkedIn API (or lack thereof), the terrible results of Twitter’s search, the vanity url issue on Facebook, etc. Hands-on experience is worth a lot when you’re trying to create a new social media business unit.
What else have you planned?
We have plenty of ideas (and too little time). Very high on our backlog is an API for developers, so that they can simply plug into our system to let us do the search, and concentrate on their own systems and GUI’s. We will specialize in further increasing the quality of search results. And we will work on building intelligence into the system, so that it will learn from earlier searches.
And of course, if you have some better ideas, feel free to hit the Feedback button on left side of the screen!
Twitter -
Subscribe -
Newsletter -
Bookmarks
| Latest, greatest and favoritest posts:
Top 200 Blogs for Developers (Q3 2009) Your Project Will Suffer from Power Laws Social Media, and Social Me |
