Business Intelligence Processing on the Base of Unstructured Information Analysis from Different Sources Including Mass Media and Internet

摘要

This article describes approaches to the business intelligent processing on the base of unstructured information analysis. The problem of reputation management of companies is considered. The review of linguistic, statistic and semantic instruments of text analysis is presented. Several examples of natural language texts analysis are discussed. The information is extracted from the natural language texts from different sources. Entities, relations and processes are built in the form of semantic network. The questions of model constructing of a subject domain are discussed. The models are presented as fragments of semantic network built on the base of the declarative instrument DECL.