| Abstrakt: |
Class of structure-independent databases nowadays is flexible data storages widely used in BI modules and information systems with a variable structure. It allows the user to store heterogeneous data, including unstructured data. Unlike relational, structureindependent databases gives user an opportunity to independently change the structure of the application and the data structure without compulsory knowledge of the query language. Data manipulation procedures, as well as structure-independent databases themselves, must meet the requirements of variability, they should be able to dynamically change not only the data values, but also the elements of the database structure: entities, attributes, etc. The methods of creating such data manipulation procedures use relational technologies that do not correspond to the capabilities of flexible databases. The paper proposes an approach that allows to synthesize flexible data manipulation procedures in structure-independent databases. The aim of developing approach is to identify and analyze heterogeneities in the existing process of procedure creation. Two types of heterogeneities were identified associated with various forms of representing the minimal structural units for constructing the process of the data insertion procedure creation, as well as semantic heterogeneities associated with the use of elements that are semantically heterogeneous to structural elements of a structure-independent database. To eliminate them, you need to use an approach based on the action designer. It allows to use actions as minimal structural units in the synthesis process, to reveal the meanings of characteristics, presenting them in the form of an action structure. The result of the paper is a developed approach which can be applied for all structureindependent databases. [ABSTRACT FROM AUTHOR] |
