Classifiers' meta-description and verification in the automation system

Abstract

Meta-description is necessary to organize, share and operate with information resources in the automation system. But standard universal meta description systems are not well suitable for classifiers. The paper analyzes classifiers’ specifics and offers appropriate meta-attributes set, based on Dublin Core standard. Meta-description includes some general attributes from this standard as well as such specific parameters, as classification and coding methods, developing, maintaining and affirming organizations, approval and implementation dates and so on. It provides the full and irredundant description for classifiers. Another point of interest for this research is a classifiers’ verification. Mistakes in classifiers’ content cause many problems for automation system personnel and software, for example, data equivocation and incorrectness, system failures and etc., and cause a system to provide fallacious solutions for its tasks. Such effects are impermissible in critical systems. Different verification aspects were examined including code length, structure and alphabet, changes by status field, code and UUID duplication and etc. All this were summarized to the set of stages for automated verification process. Its application increases classifiers’ quality and lets automation system to produce correct authentic solutions.