Image via klausonline

There should be copyright exemptions for text mining in research.

There is a fundamental shift happening now in how research is conducted and it is affecting all fields of academic endeavor. Some fields have already shifted and some are just beginning to, but the shift has a common cause, and that cause is the growing amount of research output. At a certain point, the amount of research output exceeds the ability for researchers to consume it all as it is published. In biological sciences, the shift has already begun, but the difficulties reach all the way to the (digital) humanities.

At Mendeley, we’re building tools to address this problem. Mendeley Suggest is designed to suggest relevant research to you, in effect showing you the results of searches you haven’t run yet. Searching the Mendeley catalog allows you to find papers in smarter ways than just keywords, by ranking the results according to how widely read the paper is and by showing you groups and other concepts related to the paper. At the end, though, there has to be a researcher reading the paper and using the knowledge to inform their research, and this just doesn’t scale. We need to be smarter about this. However useful these tools are, they only stem the flood, when what we should be doing is building boats.

If we want to capture and effectively use the collective knowledge discovered by the world’s academics and published at a rate of 1.5 million publications per year, we must admit that there’s a limit to how much an entirely human mediated process can handle. We have to put robots in the chain somewhere. In the assembly line of the great factory of human knowledge, we need automation. We need modernization. Unlike the manufacturing plants of the industrial revolution, both grant funders and researchers want this revolution to happen. So why isn’t it happening? It’s happening because long ago we signed away our rights to build knowledge how we want in order to support a toll-access distribution infrastructure that insists on using the same processes that worked 200 years ago to distribute today’s research. This is unacceptable. Grant funders don’t want it, researchers don’t want it, and the public has begun to demand better utilization of research funds and faster development of cures. We need to be able to mine publications, semantically link them, and have our robots build knowledge models for us. We need copyright reform. Please join me in supporting the JISC call for copyright reform & exemption for text and data mining. They want to hear from you.