A hierarchical clustering strategy for very large fuzzy databases

Accessing very large fuzzy databases is often inefficient in terms of record retrieval. Multiple probes of the database are required to obtain records that are close, but not perfect, matches. This article proposes a fuzzy database organization and clustering of records, that provides for efficient and accurate fuzzy retrieval. Additionally, a robust collection of set operators and fuzzy operators are defined. The set operators allow for the fuzzy retrieval of records based upon a cardinality constraint. The fuzzy operators are embodied in the set operators and perform the low-level fuzzy or perfect matches