Commit f0702c3e042e276cdf382400d1dd17822d0008d4
Apply https://github.com/ssbc/docs/pull/20.diff
cel committed on 3/27/2017, 1:33:16 PMParent: 8558b223f6c1a332097d438b20e64e808e4af529
Files changed
stories/design-challenge-sybil-attacks.md | changed |
stories/design-challenge-sybil-attacks.md | ||
---|---|---|
@@ -22,12 +22,12 @@ | ||
22 | 22 … | The other approach is to use social/trust networks. |
23 | 23 … | Services like Facebook or Twitter are significantly less vulnerable to spam (if you dont consider the ads spam). |
24 | 24 … | Instead of receiving personal messages from anyone one who knows your id/address on a social network, you only receive messages from other users you have "followed" or "friended". |
25 | 25 … | |
26 | -I like to call this a "Solicited Spam" system. | |
26 … | +We like to call this a "Solicited Spam" system. | |
27 | 27 … | Following someone on Twitter means you think they are human, or at least, an amusing bot. |
28 | 28 … | This is actually a pretty strong signal. |
29 | -Email doesn't an explicit social network, but the best spam filtered email systems (like gmail) can use the implicit social network in email: if I send you and email and you reply then we are "friends". | |
29 … | +Email doesn't have an explicit social network, but the best spam filtered email systems (like gmail) can use the implicit social network in email: if I send you and email and you reply then we are "friends". | |
30 | 30 … | |
31 | 31 … | Then graph analysis can be applied to the friend network - spammers may be isolated, or clustered together and can be filtered out. |
32 | 32 … | |
33 | 33 … | By building SecureScuttlebutt around a social network, preventing spam becomes easy. |
Built with git-ssb-web