E-learning creates and continues to actively develop new type of learning due to the organization of the educational process, typical for the third stage of development of the educational institutions of mankind. Creating cloud-based virtual educational enviroment is an example of creating a model for understanding the problems of analysis and design of e-learning.

The article contains the analysis of the state of information and analytical support of scientific activity under conditions of the libraries of the Republic of Belarus. A general assessment of the process of information and analytical support of science is given. The main developments and projects in the field of open access to scientific and educational resources of Belarus are described.

Analysis of tendencies in the system of education of the Republic of Belarus is presented in the article. Modern technologies and techniques, the main platforms for innovative teaching in higher education institutions are considered. An information system for ITeducation is proposed.

This article describes computer systems for teaching pronunciation and intonation which are designed specifically to evaluate and improve pronunciation while learning a foreign language. The evaluation of developed computer programs that can supplement the learning process in a classroom and provide unique prerequisites for mastering a non-native language is given: overcoming the language barrier, reducing anxiety, accessibility and individual training program. Positive factors of computer systems for learning pronunciation and intonation are noted, which support the system data with detailed and corrective forms of feedback while still meeting all the requirements of the system. The regularity of the fact that the training of a prosody of a foreign language requires a palette of intonation contours to convey the whole variety of thoughts in speech. Therefore, intonation should be studied in the context of well-structured dialogues or discourse.

A two-stage algorithm for solving the optimization task for finding the Pareto set on a given finite set of initial data N is proposed. A method is developed for finding a subset of dominated elements of the initial set N by constructing the Pareto layers on two-criterial subspaces that does not use backtracking algorithms. An iterative algorithm for constructing the Pareto set is proposed.