ABSTRACT
Formal concept analysis has recently been applied to data analysis, visualization, and knowledge extraction in various fields. Central to formal concept analysis is the notion of a formal concept that is a particular cluster in data. This chapter presents an extension of a basic setting of formal concept analysis. It allows a user to enter, along with the input data, his or her priorities regarding relative importance of attributes. Adding attribute priorities results in extraction of only those clusters in data that are compatible with the attribute priorities. The main effect is that the user is supplied with a smaller number of more relevant clusters and is thus not overwhelmed by a
13.1 Introduction ............................................................................................... 211 13.1.1 Content in Brief ............................................................................ 211 13.1.2 Introduction to Formal Concept Analysis .................................... 212
13.2 Attribute Priorities in Formal Concept Analysis ....................................... 214 13.2.1 Why We Need Attribute Priorities
in Formal Concept Analysis ........................................................ 214 13.2.2 Modeling Attribute Priorities via
Attribute-Dependency Formulas ................................................. 215 13.2.3 Some Results on Mathematical
and Computational Tractability ................................................... 216 13.2.3.1 Entailment and Its Efficient Checking ....................... 216 13.2.3.2 Complete System of Deduction Rules
for AD Formulas ........................................................ 216 13.2.3.3 Computing a Complete Nonredundant Set
of AD Formulas .......................................................... 216 13.3 Applications in Marketing ......................................................................... 217 Acknowledgment ................................................................................................... 221 References .............................................................................................................. 221
possibly large number of all formal concepts that logically exist in data. In this overview chapter, we present the approach and illustrative examples from marketing.
