Practical approach to semantics

All knowledge representation techniques have to solve the problem of semantics, i.e. how to represent the meaning of the represented knowledge. Basically, there are two possible approaches: explicit and implicit context approaches. Actually, all known knowledge representation techniques are based on the explicit context representation where the meanings of the represented concepts, properties and relations are defined by names. The explicit context approach is illustrated on the example of Linked Data concept for the structured representation of data on the Web. The implicit context representation is based on contexts defined by natural language structures (words, phrases, paragraphs, documents) and the meaning is defined using semantic categories representing semantically relevant words and phrases in different contexts. Unlike the explicit context approach, which requires an effort of specialized experts, the implicit context approach is based on various kinds of learning, where human support is only needed in supervised learning.