Wednesday, February 22, 2012

Open Science

Michael Nielsen
In the TEDx presentation "Open Science" Michael Nielsen tells several stories about successful and less successful open science projects. 

First he talks about Tim Gowers, a UK mathematician, who posed the question on his weblog if massively collaborative mathematic is possible. He posted a difficult mathematical problem on his blog, and his ideas, inviting others to post their suggestions in the comments of the blog. After a few hours people started to contribute. In 37 days, 27 people posted more than 800 comments, building on each others work, to solve the problem. 

This so-called polymath project shows, according to Nielsen, that we can use the Internet to build tools that expand our capability to solve difficult intellectual problems. For science in general, this could mean an expansion and acceleration of the way we attack a larger variety of problems. 

Another example Nielsen uses shares is the Qwiki, the Quantum Physics Wiki. This was supposed to become a repository of specialist knowledge in quantum computing. Written by the users. And although there was excitement when it was announced, no one was all that interested to contribute to it. Everyone hoped someone else would contribute. 

Many similar science wikis wave failed. Also social networks of scientists have not brought what people expected of them. As reason for these failures, Nielsen argues that young scientists have to compete for jobs, and are forced to write scientific (i.e. peer-reviewed) papers. This is valued higher than the most brilliant contributions to an open science web site. The polymath project might have had an innovative method, the result was still a scientific paper. 

The Bermuda Principles
An example from the 1990s where a different approach was successful was the human genome project. Data were openly shared on the Internet. They had similar problems at first. After a meeting in Bermuda, leading scientists discussed this problem and came up with the Bermuda Principles: 
  • Once human genetic data is taken in the lab, it should be immediately uploaded to a web site; and
  • The data should be in the public domain. 
So anyone working on human genetics, had to sign up to this and it was baked into policies of large funding institutions. And now the human genome is completely known. 

Culture change
Nielsen and other people in the 'Open Science Movement' suggest there is a need for a culture change in science so that scientists become much more motivated to share. Sharing should part of a scientists job. There should be rewards for this. Also, any publicly funded science should be open science. Everyone has a role in this, by spending at least a little bit of their time in open science projects, and be generous in crediting colleagues for their contributions in open science projects. 

Definitions definitions definitions
The suestion what exactly is open science has kept many people awake. So time for a standard! Science Commons published a set of principles for open science: 
  • Open access to literature from funded research; 
  • Access to research tools from funded research
  • Data from funded research in the public domain
  • Invest in open cyberinfrastructure
JB: So I guess this focusses on the fact that something that has been funded by the tax payers money should be accessible to the taxpayers. 
Science Commons
In their 'Introduction to Science Commons,' Wilbanks and Boyle paint the picture of a talented researcher who with all the talent, access to literature, grants, commercial tools, and all the best will in the world, would only be able to study the tip of icebergs of potential leads to study. That, combined with the argument that not everyone is that lucky to have all this access to resources, drives them to suggest a science commons, based on the creative commons principles, with 
"[..] the ambition of achieving for the world of science and data, what Creative Commons had begun to achieve for the world of culture, art and educational material: to ease unnecessary legal and technical barriers to sharing, to promote innovation, to provide easy, high quality tools that let individuals and organizations specify the terms under which they wished to share their material."
 The article further describes some of the projects the organisation started to work on to address the challenges that have been encountered in the process so far (JB: to be continued ... .) 

Creative Commons - Science
Since 2004 (CC) has been focussing their efforts to extend the Creative Commons licenses to scientific and technical research. The Web page lists several links to resources that serve different audiences in the Scientific realm with publications, repositories, and open data. 

OpenScience Project
This project is dedicated to the development developing open source scientific software. A group of scientists want to encourage a collaborative environment where data and models can be analysed and shared. 

Open Notebook Science
To prevent scientists from sitting on their data and only sharing what is convenient to them and for all kinds of other good reasons, an extreme form of open science is that where the entire primary record of a research project is made available online as it is recorded. This involves putting the researcher's notebook online together with all raw data, hence Open Notebook Science

It ennobles other scientists to compare with their own work or to build on it. This can speed up the process. It enables more effective communication. Also one can refer to the exact instances of the experiments used to support arguments. And research is being reported on, on an ongoing basis without much delay or filter. 

Arguments against
Data theft or being scooped. Difficult to publish in traditional peer reviewed journals. And there might be too much data for anyone to handle. 

Personal reflection
Not sure how it works practically, but I must say that Open Notebook Science was something completely new to me. Quite col actually that there are researchers out there doing that, no?!

No comments:

Post a Comment