Ontological Modeling of the Training Recommender System Knowledge Base

In this article, we present the results of research on the ontological modeling of a knowledge base for a training recommender system, which addresses the problem of an integrated approach to organizing the training process while considering individual user indicators. By incorporating the main components of the training process into a single recommender system, we have expanded its functionality. This comprehensive approach, which includes optimally selected exercises, a tailored diet plan to meet energy needs, and a proper recovery regime, collectively contributes to a synergistic positive effect on users’ overall health. Thus, the actual problem of providing personalized recommendations on training programs, caloric intake, and macronutrient requirements in the context of increased demand for regular physical activity was solved. The method for creating the knowledge base of the recommender system consists in ontological modeling based on semantic networks. During the knowledge structuring phase, the input and output factors critical to the outcomes of logical inference were determined. This facilitated the formation of concepts and attributes, with the individuals representing knowledge base facts. Utilizing a set of individuals with asserted relationships, SWRL-based rules were developed, enabling the semantic reasoner to infer new knowledge. Consequently, the developed ontological knowledge base model (OKBM) functions as a personalized recommender system. Results of assessing the correctness of the recommender system based on the new OKBM were obtained by testing it using different scenarios. The system evaluated users’ parameters and provided each user with an individual training program and diet recommendations focused on this program.