Fuzzy segmentation

Date of publication: September 1, 1997

Author: Erkani Keyman

Abstract:

Segmentation plays an important role in marketing research. Techniques used implementing segmentation so far only differ from each other in details. Seeking sharp boundaries between segments is the most common handicap of these techniques. In reality, there are no such boundaries. Transition from one segment to the neighbouring one is gradual rather than abrupt. One sees only tones of greys around the borderline. Classical two valued logic cannot recognise this grey gap. In this study an entirely new approach, based on fuzzy logic, is employed to fit the reality. A small test survey, based on what one might call a fuzzy questionnaire, was conducted to collect TRUE fuzzy data which in turn was used to allocate respondents to segments that are constructed through fuzzy rules.

  • PDF
  • This could also be of interest