An approach for selecting countermeasures against harmful information based on uncertainty management

Abstract

Currently, one of the big problems in the Internet is the counteraction against the spread of harmful information. The paper considers models, algorithms and a common technique for choosing measures to counter harmful information, based on an assessment of the semantic content of information objects under conditions of uncertainty. Methods of processing incomplete, contradictory and fuzzy knowledge are used. Two cases of the algorithm implementation to eliminate the uncertainties in the assessment and categorization of the semantic content of information objects are analyzed. The first case is focused on processing fuzzy data. The second case is based on using an artificial neural network. An experimental evaluation of the proposed technique have shown that the use of both cases makes it possible to eliminate uncertainties of any type and, thereby, to increase the efficiency of choosing measures to counter harmful information.