Software tools and mathematical foundations of informatics

The paper provides a brief review of the ANSI/NISO Z39.50 (ISO 23950) standard and the CIMI Profile. This standard Z39.50 describes the Application Service and specifies the Application Protocol for search and retrieval of information in databases. Z39.50 is intended for unification of network access to databases and specifies procedures and formats for a client to search and to retrieve database records. The CIMI Profile is a set of technical specifications for using ANSI/NISO Z39.50 to search and retrieve information related to our cultural heritage. Z39.50 basic services, information search and retrieval models are considered, the definitions of Abstract Database, Access Points, Database Schema, Abstract Re-cord Structure, RPN Query are determined, Attribute Sets and Tag Sets defined in Z39.50 are given, the definitions are illustrated by examples. The paper gives a brief review of the CIMI Profile. The database records search, selection and transfer specifications are considered; CIMI-1 Attribute Set, CIMI Tag Set, CIMI Schema and Abstract Record Structure are given; three levels of semantic interoperability are described; an example of Z39.50-based distributed information retrieval system is given.

The paper describes the main approaches to implementation of optimizing transformations in the SFP programming system. The algorithms of some conventional optimizations are also given.

The problem of large data array processing often arises in various scientific research and applications. This paper illustrates the ways of working with large linear data arrays by the example of processing the results of multiresolution wavelet analysis of nucleotic sequences. Some strategies of data prefetching are described. In addition, the methods of processing the nucleotic sequences and displaying the results of such processing are considered.

In this article, Ehrharts polynomials are introduced and several applicaitons to NUMA architectures are shown.

Recently, hypermedia systems become more and more popular as tools for a user-driven access to information. Adaptive hypermedia systems bring together the ideas from hypermedia systems and intelligent tutoring systems, and enable personalized access to information.

In the paper the problems of distance education support are considered, most attention is given to analysis of methods and tools of adaptive hypermedia that are used by modern educational adaptive Web-systems.

The paper is devoted to loop parallelization algorithms - one of the most effective restructuring transformations. These algorithms make use of different mathematical tools and various representations of data dependences. In the paper we describe the main loop parallelization algorithms and assess their power and limits by using the examples, as well as by comparing the "optimal" results.

The paper outlines a new programming language Zonnon being under development in the Institute of Computer Systems in Zurich. The language is aimed to be a modern alternative to a well-known Oberon language which is an evolution of Pascal and Modula-2 languages.

New approaches and technologies can greatly change the process of an interaction during a cooperative work. A special emphases is given to different ways of an anisochronous interaction which remove the problems of communication and difficulties connected with organization of meetings and telephone or electronic conferences. At present, the variety of means of anisochronous communication greatly increases. If earlier we were limited to the text form of electronic letters and bulletins, now this form of interaction becomes more flexible.

This paper describes our experience of deployment of the electronic groupware environment aimed to support collaboration on the basis of WikiWiki technology.

The paper describes the current state in constructing PROGRESS-2 - a system for fast prototyping of parallelizing compilers. The system is designed as a constructor for creating a compiler prototype whose blocks are different functional and instrumental components. Particular attention is given to the system kernel and tools for specifying a scenario of work of the compiler under construction.

The paper describes the following: the problems decided by neuronets; the methods of data pretreatment in the Significo environment; the description of standard and new architectures of neuronets supplied with the results of comparative tests; recursive and scholastic methods of teaching a neuronet; an advanced recursive method and comparative testing of methods; directions and approaches in the development of the interface part of the Significo environment.

A set of computational tools has been developed for a complete search of ciselements with the maximum use of available biological data.

A wide range of DNA analysis algorithms based on visualization of genetic information have been developed.

The algorithms have been implemented in C++ in a common object-oriented environment. Every algorithm can be used separately and as an additional filter for another algorithm's results. Implementation allows a new algorithm to be easily developed on the basis of already existing ones. The system was successfully used in several organizations.