A specific view on MSDF technology is proposed as a conceptual model of knowledge mining and processing. All levels of information processing are depicted at a higher level of abstraction and inference. The resulting structure is used for general interpretation of the Information Space that incorporates the common MSDF product-oriented model. Its examination with time lag of twelve years allows to formulate basic tendencies of information space evolution at the “zero” epistemological level. As a result, automatic MSDF is interpreted in this framework as a self-learning system. The cyclical reverse information space relations are treated in relation to the Cyberspace paradigm. Major principles of information space abstraction and fundamental axioms are formulated.