As far as we know, science building blocks are concepts which are intermingled and forms a complex fabric. We can represent science structure as a network represented by a graph. Each concept is represented as a node, linked to another by an arc. So, this set of nodes and arcs is a graph , where N are nodes and E, edges.
We argue if this network presents features like clusters and cliques very common in graphs and occurring in citation graphs or coauthoring networks. If our hypothesis is right, clusters could be mapped into sub-domains of knowledge. For example, big clusters could be identified as classical mechanics, or electromagnetism etc.
We have no guess about the structure of this supposed network. Is it a scale free network? If it does, we would have as a probability of a given node having links as .
In order to answer this and other related questions we decided:
1. To build a Jena RDF simple Physics conceptual graph using Eclipse. This graph structure is , the link could be, for example, a relation like dependsOn. The sentence, wave dependsOn frequency, could be represented as ;
2. Analyse this graph using JUNG tools.