PLAGIARISM ON WIKIPEDIA


Mon 6 Nov 06


Wikipedia-Watch.org owner, Daniel Brandt has published a report showing widespread plagiarism in Wikipedia.

A more sophisticated methodology could be used here. One idea that has just hit me would be using archive.org and the Wikipedia history function to discover phrases that appeared on other sites first. This could be all automated.

But such activity should be applauded. Whatever people may speculate about Daniel Brandt's motives, the same thing could have been done by a Wikipedian with the same positive effect: as an open system that assimilates criticism, it will learn and evolve from negative press out there. And as a fast moving encyclopedia, it can react quickly to address the problem.
COMMENTARY

by carlb on Mon 11 Feb 08
There are already plenty of obnoxious robots which purport to search new contributions to Wikipedia for signs of plagiarism, and many false-positives are being generated. From the OrphanBot that flags images as stolen based not on the image content or text but solely on the presence (or absence) of a template indicating the license, to bots that flag entire pages as plagiarism if they quote a couple of sentences of text (with attribution) from an external website, this has gotten out of control. The problem is an over-reliance on these programmes; one has to question who is the robot and who is the WP administrator here - the robot tells the admin to delete, and the admin blindly obeys. I can't see a wholly-automated approach working here. Either this has to be done by real people (not robots running unsupervised and amok) or this will produce many false-positive results and only serve to alienate legitimate contributors to the encyclopædia.

Post your comment
Your name
Your comments
Please enter the 5-letter protection code