Personalized support for users becomes very important and integral part of many information systems. It allows proposing to a user courses or news which the user may be interested in. Personalization makes information systems more user-friendly, increases their popularity, users more often come back to such systems and generally users' satisfaction increases. This paper provides comprehensive overview of recommender systems applicable for e-Learning, their architecture. It shows role of personalization for e-Learning and opportunities which can be covered by this technique. The work describes types of recommender systems and the ways they can be used in e-Learning; data-mining algorithms and methods are provided. Ways of their optimization for e-Learning environment which allow decreasing recommendation time and increasing recommendation quality are shown. А hybrid personalization system for e-Learning environment has been proposed.