Diesner, J., Pfeffer, J., & Katzmair, H. (2006). Evaluating Scientific Classification Systems by Bottom-Up Clustering of Abstracts Using Semantic Network Analysis. XXVI Sunbelt Social Network Conference, Vancouver, BC, Canada, April 24-30, 2006. Slides available upon request.
Abstract: Scientific research and development is top-down clustered by the use of international standardized classification systems (e.g. Natural Sciences, Humanities ...). By conceptualizing science and the used vocabulary of concepts as a complex emergent system it could be expected that the structure of the research area is changing permanently. This leads to the question which parts of the classification systems don’t fit any longer real world research? Which scientific areas actually now have blurred borders? Our presentation shows a way to evaluate top-down classifications of large numbers of texts and tries to characterize the European research space. The project on hand is based on semantic network analysis with AutoMap and FAS.net Software of the abstracts of about 2,500 projects funded by the European Union.