A description of the type of content is required so that the computer can collect the meaning of content. In addition, it needs defined rules in order to create the necessary connections. Taxonomies and ontologies are used for this purpose.
Another method is so-called tagging. In the case of a website, for example, it is possible to provide corresponding classifying additional information in the source code. This can be information about the language of the website as well as the topic and whether the elements used are text, image or video, etc. This makes it easier for search engines to categorize the page or its contents. This is how, for example, videos are displayed in a pure video result list.
Although the idea of the semantic web has existed for many years, it is still not widespread. One reason for this is certainly that complexity grows as more terms are connected to each other or are somehow related. In addition, there are questions regarding the degree of detail of modelling, how highly complex relationships can still be intuitively surveyed, and who ultimately decides whether the meaning of a term is correct. For this purpose, tagging systems are mainly used today as an alternative.
See also:
https://www.e-teaching.org/technik/vernetzung/semanticweb
https://de.onpage.org/wiki/Semantic_Web
http://lod-cloud.net/