KNOWLEDGE ACQUISITION FROM QUESTIONNAIRE DATA USING SIMULATED BREEDING AND INDUCTIVE LEARNING-METHODS

Marketing decision making tasks require the acquisition of efficient d ecision rules from noisy questionnaire data. Unlike popular learning;f rom-example methods, in such tasks, we must interpret the characterist ics of the data without clear features of the data nor pre-determined evaluation criteria. The problem is how domain experts get simple, eas y-re-understand, and accurate knowledge from noisy data. This paper de scribes a novel method to acquire efficient decision rules from questi onnaire data using both simulated breeding and inductive learning tech niques. The basic ideas of the method are that simulated breeding is u sed to get the effective features from the questionnaire data and that inductive learning is used to acquire simple decision rules from the data. The simulated breeding is one of the Genetic Algorithm based tec hniques to subjectively or interactively evaluate the qualities of off spring generated by genetic operations. The proposed method has been q ualitatively and quantitatively validated by a case study on consumer product questionnnaire data: the acquired rules are simpler than the r esults from the direct application of inductive learning; a domain exp ert admits that they are easy to understand; and they are at the same level on the accuracy compared with the other methods. Furthermore, we address three variations of the basic interactive version of the meth od: (i) with semiautomated GA phases, (ii) with the relatively evaluat ion phase via AHP, and (iii) with an automated multiagent learning met hod. Copyright (C) 1996 Elsevier Science Ltd