Сontext data processing for intelligent system

В статье описана библиотека программ позволяющая повысить достоверности исходных данных в контекстных приложениях. Приводится анализ применения различных методов фильтрации и сглаживания временных рядов. Оценивалась эффективность методов экстраполяции и линейного предсказания. Context-aware system is employed in a variety of contemporary information systems in order to enhance the naturalness, flexibility and convenience of user interface. Context is a mode of communication according to human senses (sight, touch, hearing, smell, and taste) or type of computer input devices (input devices equivalent to human senses: cameras (sight), haptic sensors (touch), microphones (hearing), olfactory (smell) and input devices that do not map directly to human senses: keyboard, mouse, writing tablet, motion input and others). A multimodal interface provides several distinct tools for input and output of data. Data fusion input data is much researched in the field of sensor networks where the primary motivation is to reduce communication costs by the integrationof similar data sources