Informally, a graph is set of nodes, pairs of which might be connected by edges. In a wide array of disciplines, data can be intuitively cast into this format. For example, computer networks consist of routers/computers (nodes) and the links (edges) between them. Social networks consist of individuals and their interconnections (which could be business relationships or kinship or trust, etc.) Protein interaction networks link proteins which must work together to perform some particular biological function. Ecological food webs link species with predator-prey relationships. In these and many other fields, graphs are seemingly ubiquitous. The problems of detecting abnormalities (outliers) in a given graph and of generating synthetic but realistic graphs have received considerable attention recently. Both are tightly coupled to the problem of finding the distinguishing characteristics of real-world graphs, that is, the patterns that show up frequently in such graphs and can thus be considered as marks of realism. A good generator will create graphs which match these patterns. In this paper we present gFSG, a computationally efficient algorithm for finding frequent patterns corresponding to geometric sub graphs in a large collection of geometric graphs. gFSG is able to discover geometric sub graphs that can be rotation, scaling, and translation invariant, and it can accommodate inherent errors on the coordinates of the vertices.
Priyadarshini, Sadhana and Mishra, Debahuti
"g-FSG Approach for Finding Frequent Sub Graph,"
International Journal of Computer and Communication Technology: Vol. 1
, Article 5.
Available at: https://www.interscience.in/ijcct/vol1/iss3/5