MODEL IDENTIFICATION OF PROPERTIES OF ERRORS IN DATA PROCESSING TECHNOLOGY

Abstract

The recognition problem is solved on the basis of the properties of the statistical classification structure errors, referred to as defects. Classification is performed by agglomerative cluster analysis on Euclidean metric. Experimental clustering carried out by varying the time and cost of finding and eliminating defects. We identified three major classes of defects – for accuracy, completeness and timeliness of the data. The analysis of each class of defects in their parameters, the weight value, the causes and others. In view of the identification of the properties of defects solved the problem of improving the quality of information systems.

Keywords

• recognition of the properties of the error
• the cluster analysis model
• the processing of data
• sources of error
• error recovery
• information systems