Towards a quantitative epistemology - Reconstructing the large-scale evolution of science

The evolution of scientific knowledge is directly linked to human history. Documentary and bibliographic archives such as the Web Of Science (WoS) and PubMed are fertile sources for analyzing and reconstructing this evolution. The proposed research project builds on the work of D. Chavalarias and J.-P. Cointet on the analysis of the dynamics of evolutionary corpora and the automatic construction of “phylomenetic” lattices of topics (in analogy with the family tree of species). Current tools are limited to the processing of medium-sized corpora and to non-interactive use. The expected contributions are both in the fields of Computer Science and Human and Social Sciences. Our aim is to develop high-performance solutions for generating and interacting with phylomemetic maps that exploit recent technological advances for parallelizing tasks and algorithms on complex and voluminous data. These solutions are designed and validated in collaboration with experts in philosophy and the history of science on various large-scale corpora.

Description: http://www-bd.lip6.fr/wiki/site/recherche/projets/epique/start

Software: https://gitlab.lip6.fr/equipebd/epique_prototype

Ke Li, Hubert Naacke, Bernd Amann. Exploring the Evolution of Science with Pivot Topic Graphs. In International Workshop on Big Data Visual Exploration and Analytics BigVis at EDBT 2020. Copenhague, Denmark, Mar 2020.

Ke Li, Hubert Naacke, Bernd Amann. EPIQUE: Extracting Meaningful Science Evolution Patterns from Large Document Archives. In International Conference on Extending Database Technology (EDBT demo). Copenhagen, Denmark, Mar 2020.