Questions
Is Science subjected to laws of evolution as living species are? Are there patterns in science history ?
Until recently, tackling these questions with large-scale empirical protocols was hardly possible. The ever-growing availability of data about scientific publications combined with progresses in scientific fields dynamics modeling now makes it possible to envision that these questions shall soon be answered.
Approach
As a first step, David Chavalarias and Jean-Philippe Cointet (2009) have developped a method to automatically reconstruct representations of science evolution from the large scale analysis of digital repositories : « phylomemies ». A phylomemy, by analogy with phylogenies in biology, describes the transformations of scientific fields over time: from most simple events like field emergence or decline, to more complex transformation like field merging or splitting. This poster presents, for the first time, a still partial representation of a phylomemy, related to Future and Emerging Technologies, built from the analysis of tens of millons of scientific paper metadata.
Phylomemies can also be used as high level descriptions of science evolution that, once interfaced with digital libraries, offer new ways to browse the large datasets of documents.