Abstract
For several decades, a leading paradigm of how to quantitatively assess scientific research has been the analysis of the aggregated citation information in a set of scientific publications. Although the representation of this information as a citation network has already been coined in the 1960s, it needed the systematic indexing of scientific literature to allow for impact metrics that actually made use of this network as a whole, improving on the then prevailing metrics that were almost exclusively based on the number of direct citations. However, besides focusing on the assignment of credit, the paper citation network can also be studied in terms of the proliferation of scientific ideas. Here we introduce a simple measure based on the shortest-paths in the paper's in-component or, simply speaking, on the shape and size of the wake of a paper within the citation network. Applied to a citation network containing Physical Review publications from more than a century, our approach is able to detect seminal articles which have introduced concepts of obvious importance to the further development of physics. We observe a large fraction of papers co-authored by Nobel Prize laureates in physics among the top-ranked publications.
Highlights
To a certain degree science is a system of proliferation of ideas, since even the most innovative concepts emerge as answers to open problems and are based on previously established methods: They explain seemingly incompatible data within a new theoretic framework, or they make data accessible in the first place by rendering new measurements or calculations feasible
Focusing on the picture of idea propagation rather than of credit diffusion, we here introduce an approach to ranking scientific publications in a paper citation network that is based on an intuitive interpretation of the set of all publications that have cited a paper in question directly or indirectly
Let G be a paper citation network that is composed of the aggregated information about m citations between n papers, and let i be a paper that has been published at a point in time ti
Summary
To a certain degree science is a system of proliferation of ideas, since even the most innovative concepts emerge as answers to open problems and are based (at least to some extent) on previously established methods: They explain seemingly incompatible data within a new theoretic framework, or they make data accessible in the first place by rendering new measurements or calculations feasible. Focusing on the picture of idea propagation rather than of credit diffusion, we here introduce an approach to ranking scientific publications in a paper citation network that is based on an intuitive interpretation of the set of all publications that have cited a paper in question directly or indirectly (i.e., the paper’s incomponent).
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
