In the era of rapid AI advancements and exponential data growth, web content personalization has become essential for enhancing user experience by delivering tailored information. Traditional methods, such as conceptual approaches and static ontological models, face limitations: conceptual methods lack semantic depth and struggle with domain-specific ambiguities, while ontological models, though semantically rich, are often rigid and unable to adapt to dynamic user preferences. To address these challenges, this study proposes a hybrid approach that combines the interpretability of conceptual methods with the semantic rigor of ontologies. This paper makes two key contributions. First, it introduces an ontological model that automates the identification of hierarchical and taxonomic relationships within user profiles, enabling a more accurate and dynamic representation of user interests. This enhances the efficiency of searching for relevant knowledge items and assessing their semantic relevance. Second, it presents an improved web content personalization algorithm that leverages this ontological model to better identify user interests. The algorithm integrates keyphrase extraction and ambiguity detection techniques into a unified pipeline, ensuring more precise and adaptive personalization. Experimental results demonstrate that the proposed solutions improve the accuracy of search result personalization by 5–8%, as measured by the MAP@K metric.