The Rule Extraction of Fault Classification Based on Formal Concept Analysis

An approach to extract the fault classification rules based on formal concept analysis is presented. According to the features of the real time database of the surveillance system, designed a method to transfer the records of the real database to formal context as well as to construct concept lattice, then extracted the association rules and fault classification rules. The association rule reflects the relationship among fault phenomena, and the fault classification rule indicates the fault type. A good result is gotten when the approach is used to the intelligent surveillance system of broadcasting transmitters.