Understanding Social Dynamics in Wikipedia with Revert Graph

From TrustLet, a free, collaborative project for collecting and analyzing information about trust metrics.

This paper is focused on analisys of Wikipedia social Network, and try to understand how users trust other users, in this way:

1. parse wikipedia xml history (not stub, with pages)
2. From parsed data, create a revert graph
3. From revert graph create a trust graph

[edit] Revert Graph

Revert graph is a graph created by parsing of wikipedia xml, it include:

1. wikipedia users as nodes
2. numbers of reverts between users as weight on edges

[edit] Revert

A revert is a particular type of change in the page that roll back the page to a previously version. More formally, if

a page at time t0, is in state x0, at time t1 is in state t1, at time t2 is in state t2... at time tN is in state x0

then

there is a revert between users that edit in time t1..tN-1 and users that edit page in time tN, (and you might say that user that edit in t0 trust user that edit in time tN if the two users isn't the same user) and for this reason you can say that user that edit page in time tN don't trust users that edit page in time t1..tN-1

[edit] How to use revert graph to create a trust graph

When revert graph is made, to create a trust graph we can repeat this steps for each user (named U):

1. make a set with all the users that U untrust (called Uset)
2. make a set with all the users that untrust at least one of the users in Uset (called TrustSet)
3. for all users (W) in TrustSet create an edge between U and W, and as weight insert the cardinality of the intersection between TrustSet of U and TrustSet of W

Pseudo-code Algorithm:

For U in Nodes:
  For W in Nodes:
    make a set with out-degree of U (Uset)
    make a set with out-degree of W (Wset)
    create an edge between U and W 
    put on this edge as weight the cardinality of intersection between Uset and Wset   

In this way, the trust graph is symmetric