Research On Bilateral Matching Decision-making Algorithm For Language Evaluation With Unknown Weights

With more and more bilateral matching problems appearing in various professional fields,more and more experts and scholars pay attention to bilateral matching decision theory based on language evaluation.How to deal with the language evaluation provided by matching individuals and make the final matching result better reflect the wishes of both sides is a big problem in the practical application of bilateral matching theory based on language evaluation.This paper mainly studies the bilateral matching problem based on language evaluation when matching individuals do not provide their own attribute weights.According to the maximum deviation method,the basic model of superimposing individual attribute evaluation deviation and calculating individual attribute weight vector is proposed.When dealing with deterministic language evaluation,the utility function is used to quantify the language evaluation set for single-granularity language evaluation,and the language evaluation is converted into numerical satisfaction.According to the distance formula between values,the attribute deviation is superimposed to obtain the attribute weight vectors of all matching individuals.For multi-granularity language evaluation,binary semantics is used to transform language evaluation information.According to the formula of binary semantic distance,the weight vectors of individual attributes are obtained.When dealing with uncertain language evaluation,aiming at single-granularity language evaluation,the utility function is used to transform attribute evaluation information into interval numbers.According to the Euclidean distance between interval numbers,the attribute deviation is superimposed to obtain the individual attribute weight vector.For multi-granularity language evaluation,attribute evaluation information is transformed into interval binary semantics by using binary semantics.Combining the operation formula of interval number and binary semantics,the deviation between attributes is superimposed,and the weight vector of individual attributes is obtained.Aiming at the problem that utility function can’t deal with multi-granularity language evaluation and that binary semantics can’t really reflect the individual’s psychological activities when evaluating,a new multi-granularity language evaluation processing algorithm is proposed based on utility function and binary semantics.The feasibility and validity of four algorithms for solving individual attribute weight vector and a multi-granularity language evaluation conversion algorithm are illustrated by an example.