MODELLING SEMANTIC KERNEL OF WEB RESOURCE

The article presents an attempt to describe mathematically the effect of semantic kernel a web resource on Internet. In accordance with theory marketing, product that we want sell network is characterized by following basic properties: price, time and place. other words, potential buyer wants receive given in right place at time. To satisfy this need, it necessary use classic component promotion. However, now becoming fully virtual instrument. This tool functions hypertext, video image environment. Therefore, user analyzes meaning these elements order get desired product. results projects carried out area indicate emergence new phenomenon, which reflects main promotion – core. core short annotation properties product, its location appearance. purpose both presentation object research mathematical description. It assumed formed basis natural language terms. set keywords are grouped meaning. We propose data mining approaches for clustering group method moment k-means. model based method. distance function considered as second stage content processing. At first stage, converted into web. k-means technique has significant drawbacks when modeling development idea, work shows alternative way kernel. As approach, construction clusters concept maximum flow considered. approach advantage type links overlaps result, real project, connection between level users was demonstrated over past five years.
 Keywords: kernel, keyword, k-means, max flow.