The proliferation of large databases provides both the impetus and the need for the development of algorithmic techniques for the identification and evaluation of relationships among data. This paper considers two distinct, but closely related issues: The measurement of the degree to which data satisfy a relationship and the discovery of relationships among the data in a relational database. Data associations will be described by fuzzy rules, which extend the representational capabilities of classical association rules, facilitate the construction and interpretation of rules in natural linguistic terms, and avoid unnatural boundaries in the partitioning of the attribute domains