Firefox Extension for Applied Social Networking

wanderingstan writes "Outfoxed is my masters thesis project about trust. (Nutshell overview) The extension uses a social network for personalized searching, phishing/spyware protection, file/process validation and more. It's related to del.icio.us, StumbleUpon, and those Kevin Bacon things, but goes a lot further. Mathematically, it's based on the network behavior of small world networks (pdf). Built with Javascript, Python, SQL, and XSLT. 366 testers so far, but we need the network to grow!"

Comments

Using social networks for personalization

[glinden]

Chris Anderson (of Wired and The Long Tail fame) had a great post about why social networks might not be the best way to do recommendations and personalization. An excerpt:

No matter who you are, someone you don't know has found the coolest stuff.

The sad reality is that most of my friends have rotten taste in music (I don't hold it against them), while the music recommendations I actually follow are mostly from people I've never met.

The assumption that there's a correlation between the people I like and the products I like is a flawed one.

On the one hand, you trust your friends, so things your friends clicked on might be interesting for you to know about. On the other hand, friendships are not a good predictor for recommendations since your friends often have different interests from you.

It's a cool idea, but I'm not sure how many people would bother to set this up, how often this will change the search results, whether the changes will focus your attention on the most relevant result for your search, and whether you can scale a system that accesses data on everyone in your social network on every web search.

Re:Using social networks for personalization

[rsborg]

On the one hand, you trust your friends, so things your friends clicked on might be interesting for you to know about. On the other hand, friendships are not a good predictor for recommendations since your friends often have different interests from you.

Exactly. Trust involves two aspects: competence, and compassion. Friends are often compassionate, but may not neccessarily be competant in the interests you have (ie, none of my "friends" are on in my basketball weekend group). Likewise, those guys in the bball group are not necc. guys I'd like to hang out and have beers with. But perhaps compassion can be had online, specifically for your combined interests? Some blogs have very active communities where friendships develop...

Re:Using social networks for personalization

[pizen]

What if all of the spyware companies submitted high recommendations on their links?

But those spyware companies aren't in your circle of trust so it doesn't matter what they think about their websites.

Re:Using social networks for personalization

[NickFortune]

The assumption that there's a correlation between the people I like and the products I like is a flawed one.

Well, I have a certain amount in common with my friends - that's one of the reasons they became friends. They assumption that I will like everything they like may be flawed, but the counter assumption, that there is no commonality in taste, seems equally absurd.

On the one hand, you trust your friends, so things your friends clicked on might be interesting for you to know about. On the other hand, friendships are not a good predictor for recommendations since your friends often have different interests from you.

This is about more than recommendations though. This is about whether to trust a site or product, and about what the processes on your computer do, and potentially about a lot of other stuff as well.

It's a cool idea, but I'm not sure how many people would bother to set this up, how often this will change the search results...

Isn't that the beauty of social networking apps though? If they work well, the spread; if they don't they die a horrible death. Even if this only works as a proof-of-concept, I'd say it was still tremendously valuable. We could have knowledge of trojaned downloads propagated across the net in hours, and without requiring naieve users to follow security boards either. Add in a thunderbird plugin and you could validate email links in the same way.

That's aside from the functions it shares with stumbleUpon, orkut, del.icio.us and the rest.

I do, however, take your point about setup - especially as a linux user. Still, with an established network, I can't see any reason that joining outofxed should be any more onerous than getting a gmail invite.

User != Others

[Iriel]

What I wonder is this: Yes it does seem like an interesting idea, but how many of your friends run the same software you do? I still have friends that I'm trying to convert from IE, but it's too easy for them use what's already there. I know plenty of cliques that hang together because they all like running BSD/Linux and deal with programming and such, but none of them use the same distro or the same preferences.

My bottom line is this: Look at your best friends computer. Do they have the same extensions that you do? Do they even run Firefox? The network can only be as expansive as the people that decide to jump on board.

Social networks cannot save us from dumb friends

[Flinx_ca]

If people used the brains that are supposedly inside their skulls, there would be no need for these not very useful methods of 'protection.' How many people out there would have given a thumbs up to Kazaa? My friends are great to hang out with but tend to spread the computer equivalent of STDs.

Re:sql go boom

[Jerf]

This isn't a very good idea for a host of practical reasons, mostly centering around the fact it is too simplistic.

IMHO, you are reaching for a capabilities-based model, which works out at least somewhat better in practice, though it is an open question of whether it works well enough to use. (Link leads to a group trying to build an OS on the idea, and I know it hasn't been completely smooth sailing, but I am not intimately familiar with the project.)

That should give you a springboard for further investigation into the topic, if you like. (Way too big to cover in a Slashdot post, and I am only passingly familiar with it anyhow.)

Re:who are you going to trust?

[tdvaughan]

That's no problem. You just make trust decay. With every hop away from your own directly linked network the trust metric is reduced. So I might give my Dad a trust value of 10/10 (i.e. I would trust this person with my life), but I could assign second-generation hops (those outside of my control) 80% of the trust value that Dad gives them. Allowing users to tweak their own trust decay rates will let them manage the size of their trust pool and reduce the impact of malicious users (i.e. phishers, for example).

Re:sql go boom

[Gorath99]

Unfortunately, this would just lead to:

Spyware Program [installed by] Spyware Installer [executed by]] KaZaA Installer [trusted by] User [trusted by] Root User

or

Spyware Program [executed by] ActiveX component [executed by] Internet Explorer [trusted by] Windows [trusted by] Root User

Which is exactly what's already happening.

While it would certainly be nice to have this kind of info so you can trace back where files and processes came from, it wouldn't stop malicious programs in the slightest.