Intellectual Information Technology of Analysis of Weakly-Structured Multi-Dimensional Data of Sociological Research

Abstract

To automate the process of obtaining knowledge (metadata) about the respondents of sociological surveys or questionnaires, an intelligent information technology has been developed for analyzing weakly-structured multi-dimensional medical-social data. The technology is based on the developed methods of presenting sociological data in the spaces of primary and secondary features and the neural network classification of respondents based on cluster analysis of aggregated sociological data in the space of secondary features. Approbation of the developed technology on real data of sociological surveys showed to increase the reliability of making classification decisions on the respondents’ lifestyle in comparison with the sociologist-analyst and their own definition of respondents.