"Using data gathered on Google’s takedowns over a 12-month period between April 2012 and March 2013, IPL developed a mathematical model to show how well the process is currently working and pinpoint the internal and external factors that affect consistency.To register for free access to the report, click here.
The report found that despite the volume of requests that Google receives having grown significantly (from less than two million in April 2012 to more than 18 million in March 2013), Google’s performance in terms of timeliness and accuracy has remained consistent.
As well as enabling Google to assess its performance over time, the model can be used by other companies who have to deal with similar takedown requests to perform similar analysis, and to compare this to Google and anyone else using the model.
Moreover, by modelling the relationship between the volume of requests, accuracy and timeliness, Google and others can assess the likely impact of changes to legislation – such as the imposition of a time limit for each takedown request.
Simon Morrison, Google’s EMEA Copyright Public Policy Manager, says: “How best to fight copyright infringement online while protecting freedom of expression is a thorny problem [yes, it is -- but it's a much bigger problem for copyright owners than it is for Google and "other companies who have to deal with similar takedown requests", whose headache is generally relieved at the point of takedown]. This research shows that Google has done well at balancing these important aims even as the volume of content online has increased enormously”..."
This blogger wonders how far IPL's model, and Google's takedown system, are likely to be licensed to others who are in the takedown business, and what sort of response Google might give to a request to use its software for this laudable purpose.