Research On The Retrieval Model Of Linear Algebraic Expressions

Mathematical expression retrieval is one of the key technologies to solve the problem that the existing full-text retrieval technology can not deal with the mathematical formula.While linear algebraic retrieval is an important part of constructing mathematical expression retrieval model.Due to the particularity of the form of the linear algebraic expressions,there exist many differences not only in the expression of their retrieval needs but also in the realization of index and matching methods.Therefore,when the retrieval model based on general mathematical expressions being used to retrieve the linear algebraic expressions,the results are not satisfactory.Accordingly,it is necessary to study the retrieval model for linear algebraic expressions.In this dissertation,by analyzing the characteristics of linear algebraic expressions,a retrieval model for linear algebraic expressions was proposed.On the basis of analyzing the feature information of linear algebra expression and composition of keywords,we firstly improved the mathematical expression hierarchical structure parsing algorithm,and applied it to the analytic and extraction of features of linear algebraic expression in LaTeX format,and classified the linear algebra expressions to improve the efficiency of index.Secondly,according to the basic operation of linear algebra expression,the related expressions were expanded in order to ensure the applicability and rationality of linear algebraic retrieval.Then,an index structure adapting to the characteristics of linear algebraic expression was designed.Finally,four matching algorithms corresponding to query mode were constructed.The 6352 linear algebraic expressions in linear algebra documents was used as data set to carry out experiments.The results show that the retrieval model can achieve the retrieval of linear algebra and has good performance.