System and method of multi-classification based on limited fuzzy rule in big data environment

The present invention relates to a system and a method of multi-classification based on a limited fuzzy rule in a big data environment, belonging to the big data classification field. The system comprises a fuzzy generator, a fuzzy inference machine, a basic knowledge base and a defuzzifier. The fuzzy generator performs one-to-one mapping of points determined by an input discourse domain U to a fuzzy set on the U; the basic knowledge base is formed by a plurality of fuzzy rule 'if-then' rules, the fuzzy rules comprise many types, and each type of fuzzy rule is formed by a data rule and a basicrule; the fuzzy inference machine employs the fuzzy rules to correspond the fuzzy set on the discourse domain U and a fuzzy set on an output discourse domain V on the basis of a fuzzy logic principle; and the defuzzifier performs one-to-one mapping of the fuzzy set on the V to points determined on the V. The system and the method of multi-classification based on the limited fuzzy rule in the bigdata environment greatly i