The Methods Of Rule Extraction Based On Fuzzy Linguistic Formal Decision Context

With the rapid development of big data era,the data which people come into contact with is becoming more and more complex.As the main research tool of data mining,concept lattice can extract data information effectively and help to realize intelligent knowledge representation.In daily life,the data information obtained is the linguistic values due to the vague and uncertainty of the environment sometimes.In order to avoid the lack of information in the process of linguistic value transformation,this paper uses the concept lattice to deal with the uncertainty problems with linguistic values,and the rule extraction method based on fuzzy linguistic formal decision context is studied.The main contents are as follows:1.This paper proposes a fuzzy linguistic formal context based on fuzzy linguistic values and concept lattice by defining linguistic operator,then a fuzzy linguistic concept lattice is constructed.We present a fuzzy linguistic formal decision context in view of the decision-making problems in concept lattice.After that,we study the subtle relation between concept lattice based on conditional attribute and decision attribute,and a rule extraction algorithm based on order relation under fuzzy linguistic formal decision context through the linguistic decision rules is established.We validate the effectiveness and practicability of our method via an intuitive example of traditional Chinese medicine diagnosis.2.Combined with multi-expert evaluation information,we discuss the fuzzy linguistic formal context based on multi-expert and construct a multi-expert fuzzy linguistic concept lattice.Aiming at the fuzzy linguistic formal context based on multi-expert,the distance and similarity between linguistic evaluation matrices are defined.To address the scenario that the experts' weights are unknown,we present a maximization deviation method in fuzzy linguistic formal context based on multi-expert through the distance of linguistic evaluation matrix.The linguistic aggregation operator is presented to realize the transformation between fuzzy linguistic formal context and fuzzy linguistic formal context based on multi-expert.3.In order to solve the problem of information missing under fuzzy linguistic formal context based on multi-expert,the fuzzy linguistic formal decision context based on incomplete multi-expert is proposed.Furthermore,a novel linguistic completing method using similarity and average difference is proposed to deal with the information missing problem.The fuzzy linguistic formal decision context based on incomplete multi-expert is transformedinto fuzzy linguistic formal decision context to obtain the linguistic decision rule.The corresponding confidence level and support degree of linguistic decision rule are proposed to construct a rule extraction algorithm under fuzzy linguistic decision concept so that solve the uncertainty problem in fuzzy linguistic formal decision context based on incomplete multi-expert.Finally,the rationality of the algorithm is verified by an example of multimedia teaching evaluation.