The growing complexity of manufacturing processes has intensified the demand for advanced, knowledge-driven decision support systems. This paper highlights the need to integrate three fundamental pillars: (i) smart manufacturing, (ii) emerging and disruptive technologies such as generative AI and semantic web, and (iii) multivariable and multi-domain contexts that define modern production environments. A comprehensive review of the literature identified key research efforts, challenges, and gaps related to semantic interoperability, heterogeneous data integration, and the reliability of knowledge-based decisions. Special attention is given to the role of ontologies and intelligent systems in addressing data fragmentation and fostering scalable, adaptive, and ethically aligned decision-making. Despite advances in technology, a lack of cohesive frameworks unifies these dimensions under dynamic, real-world industrial conditions. The research proposes a conceptual foundation to guide future research toward robust, flexible, and human-centred knowledge support in manufacturing. The study concludes by outlining open research questions to advance semantic integration, generative AI applications, and the development of reliable and transparent decision systems.