Wednesday, March 12, 2008

New business models for earth science data management

The value of data sharing has long been recognised and proclaimed in several manifestos and policies (e.g. Berlin Declaration, Budapest Open Access Initiative, OECD). Funding organisations most strongly support Open Access or even have initiated programs which aim to strengthen data and information infrastuctures (e.g. NSF). The importance of data archiving was also acknowledged by funding agencies and some of them have published good practice guides or data policys which aimed to convince scientists to publish their primary data in appropriate archives.

However, data management costs money. In the past, funding agencies preferred to fund the development of new systems, but they fail to ensure funding long term operation of the resulting infrastructures. Funding organisations have slowly woken up to the problem of how projects can be transformed into infrastructures but the problem is still not solved.

Even though the value of data sharing is recognised, there is little motivation for researchers to prepare their data for online access. It only causes extra work, does not add much to prestige and recognition among peers. From a researcher’s perspective, the money is better spent on further research. In this framework, policies on data sharing remain without effect.

This does not mean that researchers are unwilling. In fact, the majority is willing to share their data but in many cases are frustrated by the difficulties arising when they try to submit their data to a database. Many scientific database operators have not understood the paradigm shift in how the web works, the shift towards user generated content. I know, that mentioning "user generated content" in this context opens a can of worms. The point is, that most scientific databases, especially the publicly mandated ones, are not service oriented and simply rely on their mandate.

The funding agencies are in a dilemma. Their rules make it difficult to adapt to this rapidly evolving field. So, where is the business model to start-up independent and innovative, service oriented scientific databases? Restricted data access and paid services? This will not work because individual researchers are not able and willing to pay. Another possible avenue to obtain funding is to convince researchers of the benefits of data services and join scientific projects e.g. as subcontractors.

For this model to be successful requires motivation on both sides. User frustration needs to be avoided and technical as well as service infrastructure needs to be most up to date. Improved cooperation between data centres surely is an advantage to close own service gaps. Project specific data management networks which can share responsibilities might be a solution to satisfy user needs.

No comments: