Knowledge-Based Recommendation System for Embedded Systems Platform-Oriented Design

Abstract

Due to the increasing number of reusable solutions and proposition on the market of electronics, the search of appropriative hardware-software platforms for embedded systems design becomes more complicated and it needs more time. Therefore, the accumulation of knowledge and the development of recommendation system for solving this problem are actual scientific and practical tasks. Nowadays, recommendation system is one of the widely used online systems for making decision in different fields of human activity. In this paper, recommendation techniques are analyzed and their overall classification is presented. Different knowledge representation models are considered as well as the necessity of knowledge transformation from semantic network to other models. The developed software tool allows to represent different models of knowledge in a common language, also it contains the procedures of transformation between different forms of expert knowledge. Knowledge documentation on the basis of knowledge representation models allows to analyze the context in order to identify priority requirements and to resolve conflict requirements. Recommendation model and algorithm for helping developers in selecting hardware-software platform for embedded systems automated design are proposed. Finally, the recommendation system for hardware-software platforms is developed and presented.