ABSTRACT
We have found some propositions on this subject when having comprehensible investigation corpus such as: G. Piatesky-Shapiro developed a knowledge discovery platform-KDW and proposed the idea of “assisting the focus in initial discovery by adopting the field knowledge and finite searching” [1]; Two operations, consolidation and link formation, which complement the usual machine learning techniques that use similarity-based clustering to discover classifications, are proposed as essential components of KDD systems for certain applications [2]. Sarabjot S. Anand, based on evidence theory, developed the general frame of data mining-ESD and proposed that “users” transcendental knowledge and the known knowledge can be coupled into the process of knowledge discovery” [3]; However, none of them involved the theoretical foundation of Knowledge Discovery, not a bit of reaching the height of cognition mechanism and running system, namely there are still on the stage of scholar thoughts. In the view of philosophy, a database show the quantity of knowledge base, and knowledge base shows the quality of database. Only the research of quantity and quality, rather than one aspect only reflecting part side of nature of matters, will reflect the nature of Knowledge Discovery better. Database and knowledge base distinguish as well as connect each other, besides under certain conditions they have some counterpart relation [4,5].The specific relation between database and knowledge is exactly the start of the construction of double bases cooperating mechanism. Furthermore, the KDD algorithms mentioned only focus according to man-machine interaction, can embody the cognitive independence of
system, so they are unable to reflect novelty and validity of KDD.
