Difference between revisions of "Peer-to-Peer Accountability Enforcement"
(deleted Periscope page -- content moved to Mechanisms) |
(layout tweaks) |
||
Line 1: | Line 1: | ||
==About== | ==About== | ||
[[Peer-to-Peer Accountability Enforcement]] is a methodology for sharply reducing the problem of posting content in bad faith (including both outright verbal abuse as well as abuses that are harder to spot, such as {{l/ip|sea-lioning}}) by allowing users to collectively delegate other trusted users to rate comments and commenters as to their credibility and appropriateness. It generally increases per-user accountability for abuse, but with the source of that accountability being other users rather than a central authority (with all the bottlenecking and [[power-concentration]] that implies). | [[Peer-to-Peer Accountability Enforcement]] is a methodology for sharply reducing the problem of posting content in bad faith (including both outright verbal abuse as well as abuses that are harder to spot, such as {{l/ip|sea-lioning}}) by allowing users to collectively delegate other trusted users to rate comments and commenters as to their credibility and appropriateness. It generally increases per-user accountability for abuse, but with the source of that accountability being other users rather than a central authority (with all the bottlenecking and [[power-concentration]] that implies). | ||
− | + | ==Pages== | |
− | + | <big> | |
− | * | + | * '''{{l/sub|purpose}}''' - this needs to be a bit more general |
− | * | + | * '''{{l/sub|mechanism}}''' - the quasi-technical details |
+ | </big> | ||
==Notes== | ==Notes== | ||
Things that credibility management ''should'' be able to defeat or at least control: | Things that credibility management ''should'' be able to defeat or at least control: |
Latest revision as of 12:25, 3 May 2021
About
Peer-to-Peer Accountability Enforcement is a methodology for sharply reducing the problem of posting content in bad faith (including both outright verbal abuse as well as abuses that are harder to spot, such as sea-lioning) by allowing users to collectively delegate other trusted users to rate comments and commenters as to their credibility and appropriateness. It generally increases per-user accountability for abuse, but with the source of that accountability being other users rather than a central authority (with all the bottlenecking and power-concentration that implies).
Pages
Notes
Things that credibility management should be able to defeat or at least control:
- sea-lioning (see Issuepedia): appears civil and polite on the surface, so may be difficult to judge without understanding the full context
- brigading -- though it may take a combination of credibility management and debate mapping:
- evaporative cooling:
- click-farming ...except I'm not understanding the value of having fake followers:
- 2015-04-20 How Click Farms Have Inflated Social Media Currency
- private discussion here
- 2015-04-20 How Click Farms Have Inflated Social Media Currency
- online harassment
Credibility management is beginning to look potentially useful for rating subjective quality of aesthetic works. Some discussion of that application is here:
- 2014-06-20 Content rating, moderation, and ranking systems: some non-brief thoughts (Edward Morbius).
- Related: 2014-09-21 Specifying a Universal Online Media Payment Syndication System
- which was a sequel to: 2014-01-08 A Modest Proposal: Universal Online Media Payment Syndication
- Related: 2014-09-21 Specifying a Universal Online Media Payment Syndication System
- 2012-02-08 Tribler Makes BitTorrent Impossible to Shut Down (via) "Where most torrent sites have a team of moderators to delete viruses, malware and fake files, Tribler uses crowd-sourcing to keep the network clean. Content is verified by user generated “channels”, which can be “liked” by others. When more people like a channel, the associated torrents get a boost in the search results."
- 2011-02-05 What is Quora's algorithm/formula for determining the ordering/ranking of answers on a question?: this is a similar concept on the surface, but lacks some important elements:
- no proxying/layering -- all ratings are direct
- no personalized credibility ratings (PCRs)
- minimal granularity, i.e. only two possible values (-1/+1) for each ranking