Multidirectional transformations and Web-representations of differentlystrtuctured information

During the last years the problem of multiple usage of the same data becomes still more actual; we need to process and represent data in various formats, basing on the data that are entered once; moreover, these data may also be differently structured. It is used both in work with local databases and with Internet resources, in control and information intranet systems.
The purpose of this research is to optimally fill, transform and, what is even more important, to send and output differently structured information based on such data sets. The most actual are the problems of selection of optimal data representation formats, especially for the cases of data of big size, data of variable structure, incomplete data, as well as building instruments for their transformation, including output for Web browsers. The principal requirements, difficulties and ways of solving the given problem are studied on typical examples of “institute” and “archive”.
There are several useful approaches to better organize work and solve the problem set. First of all, it is usage of file system for data organization, usage of descriptors for information blocks on the level of folders, including those that define the structure of information that is placed in the current block; it allows to correctly select and represent the information, properly connect it with information form other blocks with the same or other structure and contents. Usage of hard and soft file links was useful in operating systems of Linux family, but not so successful for MS Windows.
The principle of selection of metainformation from archives, with subsequent interchange of only such metainformation between servers was very useful; these are metadata about presence of some information in the archive on the given server, brief or full list of information items on some fields; with semiautomatic renewal of such information.
The data representation implementation based on CSS mix was also rather useful. Inclusion of information and metainformation in simplified languages like YAML and JSON also helped to improve flexibility and speed of information selection and representation system.
In general, the economy of development time in typical cases was about 25–30% of the traditional one; but is actual only for “middle” size systems; for “small” and “big” size systems additional study is necessary. We should also more accurately define ways of efficiency measurement and perform them for “big” systems.