Half a croissant, on a plate, with a sign in front of it saying '50c'
h a l f b a k e r y
You want a piece of this?

idea: add, search, annotate, link, view, overview, recent, by name, best, random

meta: news, help, about, links, report a problem

account: browse anonymously, or get an account and write.

user:
pass:
register,


                                     

moral value competition

doing the markov-chain-thing on court rulings
  (+1, -5)
(+1, -5)
  [vote for,
against]

court rulings normally have a justification tacked on them which states which paragraphs or moral values are more important than others.

so let's do the markov-chain-thing on those documents and create a list of moral values from those who win most often against others to those who lose most often.

i do not really know what to do in case of circle dependencies, but the chosen legal system probably is fucked up anyway if they occur.

erlehmann, Jan 11 2008

Algorithm + Discussion group = AI Algorithm + Discussion group = AI
Kind of reminds me of this. [zen_tom, Jan 11 2008]

[link]






       court shoes?
po, Jan 11 2008
  

       Re first para; would you please elaborate. Which courts?
angel, Jan 11 2008
  

       [angel], i don't understand your question. would you please re-phrase it ?
erlehmann, Jan 11 2008
  

       translation re first para would you please elaborate which courts
ConsulFlaminicus, Jan 11 2008
  

       Re-phrase: you say that "court rulings normally have a justification tacked on them which states which paragraphs or moral values are more important than others". Would you please elaborate as to which courts issue rulings with such justifications.
angel, Jan 11 2008
  

       Can we get a category check, please?
phoenix, Jan 11 2008
  

       [admin] Phoenix - done.
zen_tom, Jan 11 2008
  

       category check done, sorry.
erlehmann, Jan 11 2008
  

       Ahh good. I'm glad this idea came up, I remember looking at the Markov Chain section of wikipedia some time back and coming away completely none the wiser - any chance of a layman's explanation?
zen_tom, Jan 11 2008
  

       dataset:   

       A > B > C > D > E > F   

       A > C > E > B > D > F   

       A > B > C > E > D > F   

       C > A > B > D > F > E   

       a simple markov chain program now finds for B:   

       2x A > B > C   

       1x A > B > D   

       1x E > B > D   

       i now score those short chains based on how often they occur. therefore A > B scores 2, E > B scores 1 and everything else scores 0. similarly, B > D scores 2, B > C scores 1 and everything else scores 0.   

       now i concatenate several 3-item-chains and determinate their cumulative score. in this example, A > B > D has a cumulative score of 4 and wins over, say A > B > C, which is only worth 3 credits.   

       ¿ am i {understandin'; explainin'} it 'rong ?
erlehmann, Jan 11 2008
  

       Hmmm, I'm not sure - shouldn't A>B score 4 (1 of those being via an intermediary)
zen_tom, Jan 11 2008
  

       i don't think so, you are probably thinking of concordet voting. since we probably have intransitive relationships here, it is contrary to our purpose to evaluate groups bigger than three items. (edited, wrong wording, srsly)
erlehmann, Jan 11 2008
  

       A > B > C > D > E > F *
A > C > E > B > D > F
A > B > C > E > D > F *
C > A > B > D > F > E *
  

       OK, so if intermediaries don't count, shouldn't the 3 *'d lines count towards the A>B score?   

       Anyway, I get the general idea. It kind of reminds me of another one here [linked].
zen_tom, Jan 11 2008
  

       It reminds me of how LZW encoding works.
Ian Tindale, Jan 11 2008
  

       You mean circle dependencies like "paper beats rock which beats scissors which be paper again"   

       or do you mean like where criminals keep an eye on which laws are moral and which ones are okay to break in order to justify the crime?
quantum_flux, Jan 25 2008
  

       // or do you mean like where criminals keep an eye on which laws are moral and which ones are okay to break in order to justify the crime? // cannot comprehend; elaborate on that.
erlehmann, Jan 25 2008
  

       It's the old "paper stole rock because rock was about to bash scissors in" argument.   

       Obviously stealing rock was the more moral thing to do than bashing scissors in, so paper was therefore morally justified. However, bashing scissors in would have also been a good thing for rock to do because scissors was about to cut paper into pieces. This is known as the circle of morality.   

       Also, there is the anti-circle of morality as well. This is where a criminal decides against committing a crime and that leads to another crime occuring which ensures their survival. Paper didn't have to steal rock, and paper would have been much safer off if rock had actually bashed scissors in. AND THEN paper could have justifiably enslaved rock for being immoral and would have become the king of the world!   

       For you see, paper waited until it became moral to enslave rock.... he waited till there was no more retribution from scissors to contend with and then stole and enslaved rock for committing an unjust crime against scissors!
quantum_flux, Jan 27 2008
  
      
[annotate]
  


 

back: main index

business  computer  culture  fashion  food  halfbakery  home  other  product  public  science  sport  vehicle