We are starting to see Semantic Web technologies being used in our research papers and data. The next revolution will be publishing knowledge at the speed of thought.
In September 2007, Wired Magazine included an article by Thomas Goetz
entitled \Freeing the Dark Data of Failed Scienti c Experiments" [
Research|all research|has value. That value is independent of any grant awarded, tenure attained, or publisher's acceptance. The value lies in learning. Knowledge, truth, and the pursuit of understanding something previously not understood are the goals and purposes of research. Nowhere in any de nition of the word research have we read that research only has value if it produces cherry-picked outcomes that make administrators, editors, grantors, and publishers happy. And yet, due to the undeniable bias towards \positive" results|or what have been deemed as such|much of the rest of all research is sentenced to a ling-cabinet death-by-obscurity. It does not have to be this way in today's scienti c technology revolution.
Incentives are the key here. Because future job prospects, tenure, and public acclaim all tend to revolve around journal publication, why would a researcher want to take the time to go back, write out, format, narrate, and then attempt to publish negative results that are likely to end up in unpretentious journals or on personal blogs that often are not recognized as of much worth by the research community, employers, or funding agencies? That's a lot of e ort and time for something that will essentially provide no career or nancial advantages for the researcher.
That leaves two things we can do to bring dark data to light: one, change the incentives involved in research to better re ect the value of negative or inconclusive research results, and two, remove the barrier of time and e ort that keeps a researcher from sharing this dark data in the rst place. Ideally, these two changes would occur simultaneously, but funding agencies and institutional policies tend to be incredibly slow to change. Social expectations are often equally or even more sluggish to evolve. So we have a choice: do we sit around and wait until publishers and institution administrators change incentives to publish dark data? Or, do we nd ways to make it worth sharing this dark data now?
If the time and e ort barrier were removed from sharing noncon rmatory
research, we believe that many researchers would gladly share those results even
if there were no external incentives. Open notebook science gives us a good
starting place, but we need to overcome the problem of data deluge [
The processes of researching and publishing would become distinct and separate from presenting the research results to a certain audience in, for example, a paper. If the researcher wanted to write and submit a paper based on those noncon rmatory results, there would be nothing to prevent that. In fact, having the entire research process preserved would make writing up such journal articles easier anyway. Other presentation formats could be created, such as video or interactive media, without worrying that some important parameter or intricate concept might be left out. But the biggest game changer is that all of that dark data would come to light, and that the paper might not always be necessary.