Directed by the Knowledge Discovery Schema, this phase performs all the pre-mining functions for selection, cleaning and transformation on the growing database in an incremental manner[12]. Knowledge is accumulated from the growing database in the form of Knowledge Concentrates which are used in the mining phase. In this phase, there is absolutely no interaction with the user. Accumulation is performed automatically & periodically (frequency is specified by the user in the knowledge discovery schema) till the user drops the schema. Crucial Issues at this level are the interaction between the database and the knowledge discovery process to access the data records and the maintenance of knowledge concentrates on the disk.