Research On The Correlation Score Between Literatures Based On Improvement Of Approximate Method For The Expected Value

This paper uses the expected value of the degree of association as a starting point and gives a method to approximate the degree of association of the paper.For a given independent variable value,the expected value of the degree of relevance of the paper can be used to approximate the degree of relevance of the paper,but the true expected value is not known,so first of all,the expected value of the degree of relevance should be approximated.Survey literature and empirical equations can be used to calculate the initial approximation of the expected value of the degree of correlation,but this approximation is to construct a single approximation function from the survey literature.If the survey literature can be used to construct a piecewise approximation function of the expected value,then there may be a better approximation of the expected value.When constructing a piecewise function,the commonly used method of constructing a piecewise approximate function cannot be used directly.Two changes must be made to the commonly used method.One change is to replace the derivatives in the boundary conditions with their approximate values,and the second change is to replace the true expected values with the preliminary approximate values of the expected values.If the commonly used boundary conditions are used,then the changed construction method,its solution exists and is unique.After finding the piecewise function,the approximate effect of the piecewise function on the expected value should be analyzed,and it can be used if it is valid.When analyzing the effect of the piecewise function,it can be analyzed according to each value of the independent variable and from the overall perspective of the survey literature.When analyzing by each value,the confidence interval of the expected value can be used to analyze the relative error of the piecewise function value and the true expected value.According to the size of the error,it is possible to analyze whether it is effective to approximate the true expected value and the degree of correlation with the piecewise function.From the overall analysis of the survey paper,it is no longer necessary to test the effect of using a piecewise function to approximate the expected value.It can be analyzed from the correlation between the number of co-citations and the degree of association.There is at least no significant difference overall,then the value of the piecewise function can be used to approximate the degree of correlation,and this approximation function is effective in the overall survey literature.