Social Network Aggregators: Automatically Measuring The Importance Of Events

Posted: August 22nd, 2007 | Author: Ben | Filed under: Interactivity, Thoughts, Web Development |

Here’s an idea I’ve had and will never do anything about so I thought I’d put it out there and see what people think.

My problem is that on my Facebook (this idea could apply to other sites but in this post I’m just going to use Facebook) I have lots of people listed as friends. I have a few close friends, then I have friends who I know but who aren’t close - workmates, friends of friends and so on. Then the bulk is made up of the people who I haven’t seen in years, but for some reason are still my friend. The result is too much information for me to manage that isn’t effectively organised. Sure, Facebook lets me define who I want to hear more about and who I want to hear less about, but this is a manual control and I think it should be able to work this out by itself.

How can a system like Facebook work out which is more important to me: wedding photos from someone I went to school with; knowing that my flatmate ‘is bored at work’; or photos of my workmate’s new dog?

Step 1: How much do I care about the person?
Facebook could work this out from the amount of correspondence between my friends and myself on the site. Then there’s the number of photos we appear in together and the groups or networks we’re both members of. Although granted, my close friends and I don’t speak exclusively through Facebook, so some manual intervention may be necessary. But for the most-part my degree of ‘closeness’ with the friends I have on Facebook could be calculated.

Step 2: How important is the news to others?
First of all the relative importance of the news for the author could be calculated by how often the author posts news. By taking the ‘cry wolf’ concept; if someone posts news every day it is possibly less important than someone else who has posted an item for the first time in three months.

Step 3: How popular is the news to other people who know this person as well as I do?
If someone else has clicked on a news item for a mutual friend we both went to school with, it’s possibly also interesting to me. If I post an item that only my close friends have shown an interest in, it probably won’t be of interest to someone I worked with years ago and don’t speak to any longer.

Existing Features
Facebook already allows you to select the kind of news you want to see more of. So for example, if you happen to like seeing new photos from your friends, it will offer these up. If you don’t like seeing people’s status changes you can see less of them. You can also select the people you’re more interested in, as well the people you’re not interested in - although this is done manually: Facebook can’t determine how ‘interested’ you are in each of your friends by your dealings with them through the site.

Opening It Up
Of course, a lot of people are on more than one social network. A decent aggregator would sit outside of Facebook and MySpace, watching and analysing everyone’s reactions to news.

Potentially, the system could even take on StumbleUpon to monitor more than social network activity. If several people you know visit a certain website, or a news story or video, you could be alerted to the same content based on how close you are to the friends who also visited that page.

It’s a matter of managing mass-information. As more people become proactive in creating web content, there should be a way of managing the publicising of this content to the right people. Monitoring social interactions does wreak of an Owellian telescreen; but there should be a way of creating a system that would benefit its users rather than its administrators.

---