Study On Similarity Ranking Of The Retrieval Results Of Linear Algebraic Expressions

Linear algebra is an important part of modern mathematics.It is widely used in natural science and social science.With the deepening of linear algebra research,the mathematical formulas retrieve requirements for linear algebraic formulas is increasing,the existing retrieval technology and search engine theory and methods are facing challenges.Compared with the general mathematical formulas,the linear algebra formulas have more complex structures and richer syntax and semantics,so it is more difficult to retrieve them and sort the results in a reasonable order.And aiming at this problem,we study the ranking of retrieval results in linear algebra formulas retrieval model.Firstly,on the basis of defining linear algebraic query mode,the characteristics of linear algebraic structure,symbol,grammar and semantics are summarized by analyzing the characteristics of linear algebraic methodsSecondly,the fuzzy similarity evaluation attributes and their corresponding membership functions and the similarity functions of linear algebra formulas were defined from global and local aspects by using the knowledge of hesitant fuzzy set.Finally,the similarity computation between linear algebraic equations is converted into similarity computation between corresponding hesitant fuzzy sets,and the ranking of linear algebraic retrieval results based on similarity is achieved.Through the experiment of the 6352 formulas,it is proved that the method can realize the ordered output of the retrieval results of linear algebra formula retrieval system.