This paper presents an overview of data mining techniques in the context of of the deductiveinductive cycle model. The claim made in this paper is that a combination of both of these modes of inference in one model o ers adavantages over the exclusive use of either induction or deduction. After an introduction of this model the paper describes an architecture of the Knowledge Miner environment and its relation to the LDL++ deductive database technology. In particular, it de nes the notion of a meta query and shows how this concept can be used to take advantage of the full power of a deductive data base system. The paper describes a number of inductive methods that can be used within this framework and describes some data mining experiments performed with the Knowledge Miner.