Bitcoin Forum

Bitcoin has a reputation of allowing anonymous payments, but people with more Bitcoin expertise know that, without special 'bitcoin laundry' tools, you are far from anonymous when using Bitcoin.

The idea:
Wouldn't it be cool to have a website that allows visitors to do data mining on the block chain? I imagine it could be something like a fuzzy logic algorithm that assigns a probability to each pair of addresses on whether they belong to the same person / wallet, and then show transactions between those address clusters, instead of between individual addresses. Since people will often repeatedly do business with the same person/organization, this could also be used to reinforce the address matching.

On top of that, there could be a wiki-like system, where these 'identities in the block chain' are connected to real-world identities. For instance, after withdrawing bitcoins from an exchange service, I know a Bitcoin address of that exchange service. I could enter that information in the wiki, and then find out what other addresses belong to the exchange service, how many bitcoins they probably hold in total, etc..

Why is this cool?
First of all, this would be a nice demonstration to law enforcement people, to show them that Bitcoin does not necessarily remove their ability to 'follow the money'. This could help to get political support for Bitcoin.

Secondly, this could actually improve anonymity: people can actually see on the website whether their anonymity has been compromised, and they can take measures as soon as they notice.

People who want to compromise anonymity will do datamining on the block chain anyway; by making the best datamining tools available to the general public, at least we make sure nobody has an advantage over other people, and we increase visibility of what is visible and what is not.

i think this is a good idea.  why not flush out the true anonymity capabilities of Bitcoin?  if not you, someone else will eventually do it.  i'd be interested myself in being able to probe the limits of such a system and would use it regularly i'm sure.

this is a fuzzy area that i'm sure alot of ppl would be interested in.

It would sure help with showing a "good faith" effort which is all law enforcement is looking for these days. Even megaupload could have done that and they wouldn't be sitting in jail right now.

I think this is an excellent idea.  Watching.

watching, interesting idea

i guess the only problem with this is that no one would want to contribute to a wiki if that address could then lead back to themselves.

and then you'd probably have ppl entering in real addresses with bogus ownerships to obfuscate the linkages or even to target enemies.  it could turn into a real mess.

Yes.

If this were to be done, there'd be no point to the wiki. Just allow people to explore the interconnections between addresses (they can already do this with blockexplorer, it's just pretty slow.)

They'd be responsible for determining any public ID connected to a specific address themselves, since this is where things could get messy and unprovable anyway. (Just because Jane puts an address on a website and says its hers doesn't mean she's the one who owns the private key, or even that the private key to it exists.)

Would it be possible to use Google engine to search web to make connection between Bitcoin addresses and people?


IMO, If it's possible to create such tool then it would be a good thing because it would make people aware of how much personal information they share.

It could also give People the possiblitity to rate others "Identities/adresses" on a scale from 1-10 on how cool their behaviour was in doing business..

of course bad guys will change their adresses but the good guys could prove their reputation through their control of their private key..of course somehow we`ll soon have an reputation market :)

dunno but i read the sequel to deamon yesterday and this somehow goes that direction ..SmartMob anyone ^^

I mentioned something like this on another thread earlier today and dubbed it, "4 Square on Rails". If users desire, they can create a vanity address for themselves and use Firstbits's shortened version of the same as their own personal address. I have another idea centered around this that I'll be expressing on one of the restaurant related threads, unless I feel it'll warrant its own thread.

~Bruno~

This is a really great idea! Here's an idea I have that could be a nice starting point:

-If 2 addresses are used as the input in the same transaction, we can be sure they belong to the same wallet, hence the same user
-If the 2nd of those 2 addresses is again used in another transaction as an input, together with a 3rd address as input, we can be sure all those 3 addresses belong to the same wallet, hence the same user, even though the 1st and 3rd address were never used together.
-And so on and so forth. Just check for every additional address you can link to the same/wallet person, if that address is also linked to more addresses.

I'm wondering how effective this would be, and how many clusters and more importantly how big the clusters of addresses would be that we can find. I'm fairly sure this would be a good starting point, after that it might be a good idea to see if it is possible to correlate certain money flows, to bind separate clusters together in bigger clusters if analysis shows a high probability of them belonging to the same person/wallet. I'm not sure how hard this would be to implement. I think it shouldn't be too hard, but my programming is way to rusty to pull this off. Anyone willing to give this a go?

Good luck! :)

I feel we are missing out on major markets if we keep the anonymity aspect alive. If we simply rid the anonymous aspect, and embrace the transparency features, we're apt to go a lot further and quicker. And yes, we could have both, but I feel by pushing the latter, we'll be golden.

~Bruno~

BIP 16 can go along way toward allowing the option of assured identity if a credit bureau is used as a signatory on a transaction.

Anonymity is necessary because all transaction can be seen by everyone. Would you use a bank that allows anyone to log on to their web site and see any account by user name and everyone that user has received or sent money to?

+1

and this is why we should test the system to determine its resilience in providing that anonymity.

I may have to start a new thread to clarify what I'm trying to relay. We live in a look-at-me society, and what I'm thinking about is that the blockchain would be a perfect place for people to show others what they've paid for--on purpose. I can easily and quickly come up with a couple dozen nice examples, and given just a bit of more time, I could provide thousands, albeit only about 500 or so would be viable.

~Bruno~

this is fine if any individual chooses to reveal that info.  but in general, each Bitcoin address is appropriately viewed as a bank acct to which the owner has the right to privacy/anonymity.

The entire marketing plan(s) is to show those transactions, not hide them, with the ultimate goal of bringing further awareness to Bitcoin.

I am glad to see that not everyone is obsessed with anonymity. There is so much more to Bitcoin, just like there is so much more to education and grades than keeping your grades secret from others.

Not if there were a Handicapper General who penalizes people with good grades. I'll be glad to see this tested, though, may not be so pleased with the results, but we'll be better for it.

The original poster is obsessed with anonymity  ::)
He just thinks this would be a great tool for people to check their anonymity. Just like port scanners and exploit frameworks are great tools to check the security of web servers.

I also like the comment of Phinnaeus about the "look-at-me society". People do not always choose for anonymity: sometimes publicity is more attractive. However, I believe this should always be a choice. And I don't think this datamining-idea is the right thing for showing that a certain transaction belongs to you. It is much easier to do that reliably with a sort of signed message system, maybe with a sort of Public Key Infrastructure for identity verification.

This, to address the bolded phrase above: