Dynamic Network of Concepts from Web-Publications


The network, the nodes of which are concepts (people's names, companies' names, etc.), extracted from web-publications, is considered. A working algorithm of extracting such concepts is presented. Edges of the network under consideration refer to the reference frequency which depends on the fact how many times the concepts, which correspond to the nodes, are mentioned in the same documents.

Web-documents being published within a period of time together form an information flow, which defines the dynamics of the network studied. The phenomenon of its structure stability, when the number of web-publications, constituting its formation bases, increases, is discussed.

Autors: D. V. Lande, A. A. Snarskii

Source: arXiv:0806.1439v1