By Pete Meyer, Ads Quality Statistics Team

Google was a sponsor of this year’s international user conference in Dortmund, Germany for the R statistical programming environment. In addition to sessions focusing on statistical modeling in various areas of applications, the conference included a number of sessions focusing on user interfaces for R and producing high quality output for reports and browsers. There were sessions on high performance computing, including parallelized approaches to R, and new routines for handling massive objects in response to the increasing size and complexity of datasets. Google was among a number of companies who showed examples of how R can be used in production environments as a statistical tool to better understand collected data. We presented some examples (PDF, 348 KB) of how HTML and email output and database access can be integrated into R’s statistical environment. The goal is to provide new insights in a familiar format for those who oversee the human evaluation component of Google’s advertising quality efforts, without having to generate intermediate datasets.

We’re a member of the R community as both users and contributors via Google Summer of Code™. We see the R community as an ally in our mission to organize the world’s information and make it universally accessible and useful, and we are glad to be able to support their efforts as a sponsor.