If you are interested in creating your own “Maps of Science”, you might find some useful informations in the “BiblioTools” section of this website…
Research at the ENS de Lyon
I have opened a blog to share ongoing work dealing with representations of bibliometric informations (“maps of science”). I specifically use the ∼ 7500 articles with an “ENS Lyon” adress gathered from Web of Science. Your comments are welcome!
Scientometric Map of the ENS Lyon. This heterogeneous network shows the authors, keywords, references and institutions displayed by at least 10 of the ~ 8500 articles in the database we extracted from the Web of Science. Two nodes are closer to each other (and have a stronger link) if they are often used in the same articles. Inset : the global map shows the overall structure in distinct “clusters” corresponding to the different teams of the ENS labs. Detail : zoom on the central part. Notice in particular the specificity of the “Joliot Curie” biophysics lab, linking a team of physicists to a team of biologists.
We have created several maps presenting different kinds of scientific community we identified inside the ENS de Lyon (communities of articles sharing references, communities of co-authors, communities of keywords co-occurring in the same publications, … ). These maps are displayed inside the ENS buildings to spark off some comments from the researchers. For more details on this research, refer to the following:
Collaborators:
- Pablo Jensen (IXXI, ENS Lyon)
Quantitative History of Complex Systems Sciences
We have gathered a database of around 200.000 records linked to the complex systems scientific domain. Each record consists of the authors, keywords, title, journal and references of the article, as gathered from Web of Science. These articles come mainly from four science areas, namely physics, biology, computer and engineering sciences.
All the articles published during a given period defining nodes of a network, we chose to study the relationship between two publications with the notion of bibliographic coupling, ie, to define links between articles sharing some references. These links implicitly define communities of articles sharing some common grounds. These communities are first detected thanks to the clustering algorithm developed by the Louvain team based on a maximization of the modularity function and then visualized thanks to Gephi – a graph visualization software.
Since our data contains records from 1950 to 2009, we are able to detect several scientific communities and to analyze their dynamics, how they appear, grow or shrink and sometimes disappear. On a more microscopic scale, we can also follow how a reference starts to be cited by scientists from different communities (ie to become an essential link between different communities), which provides an essential interdisciplinary ingredient for the growth and stability of the field of complex systems.
The goals of this work are
- to develop useful quantitative tools for analyzing huge bibliometric database – more specifically here, the relations between different communities
- more particularly, to develop tools for analyzing and visualizing the evolution of dynamic networks
- to study how complex systems has emerged as an independent established research topic
A study on the notion of complexity as we apprehend it at the IXXI, and compared to french philosopher Edgar Morin’s view on complexity, lead to an article published in Hermes, a french magazine specialized on information and communication sciences. Refer to the following:
⇒ Hermes 2011 (proof reading version)
Collaborators:
- Pablo Jensen (IXXI, ENS Lyon)
- Guillaume Beslon (IXXI, LIRIS, INSA-Lyon)
- Sara Franceschelli (IXXI, ENS LSH)
- Eric fleury, Qinna Wang (IXXI, INRIA, ENS Lyon)
- Céline Robardet (INSA Lyon, LIRIS)
- Jean-Baptiste Rouquier (ISC-PIF)