Measuring Conflict on Wikipedia

Topics

This week’s assignments will guide you through the following topics:

Please read the following:

Edit Wars on Wikipedia, Sections I and II.
Wikipedia Policy on Edit Wars
Quartz. “Wiki Wars: Inside the increasingly nasty battle for Wikipedia’s soul.” [Link]

On the topic of edit session data, you may find this article’s careful usage of this data type interesting:

R. Stuart Geiger and Aaron Halfaker. “Using Edit Sessions to Measure Participation in Wikipedia.” CSCW 2013. [Link]

Complete the following tasks:

Read the companion webpage for the Edit Wars paper and download/analyze the following contents: WikiWarMonitor. Pay careful attention to:
- ‘Light-dump’ data, the intermediate data the authors use to identify edit wars.
- The script that transforms wikipedia dump data to light dump data.
- The script that computes the M-Statistic.
There are two articles contained in two files in the location /teams/DSC180A_FA20_A00/b03onlinecommunities/testdata. One of the files is the wikipedia edit dump format, while the other is in light-dump format. Look at the contents and identify the reversions in the light-dump data with the actual change in text in the XML (you may find it useful to look at the articles in the browser, as well). Write code that translates wikipedia edit data into light dump data. Hint: you can use hashing to tell if contents of two articles are exactly the same.

Answer the following questions:

How to the authors of Edit Wars define ‘contraversiality’ in wikipedia edits? Give two answers:
1. an answer that is understandable to a general audience,
2. an answer that is precise enough to translate into code.
Give an obvservation you found interesting in the wikipedia policy on edit wars. Try to give an observation relevant to the Edit Wars paper.
In the test data, how often does a reversion (as defined in the light-dump data) have the word ‘revert’ in the ‘comment’ field of the XML?