EFFICIENT REGULAR GRAMMATICAL INFERENCE TECHNIQUES BY THE USE OF PARTIAL SIMILARITIES AND THEIR LOGICAL RELATIONSHIPS

New techniques for regular grammatical inference are proposed by the use of partial similarities among sample sentences and their logical relationships. These have an advantage of making clear the characteristic features of resultant automata to be inferred. The inference techniques including previous ones are classified from the points of view of the fidelity to samples and the efficiency of inference. It is also shown that the proposed methods can be efficient according to these two criteria.