A Mathematical Expression Retrieval Model Based On Inter-relevant Successive Tree

Mathematical expressions include various kinds of elements such as numbers,operators,letters,and other symbols coupled with their complex and diverse structure,which makes it more difficult to realize their indexing and retrieval compared with the traditional full-text retrieval technology.In this thesis,according to the analysis and summary of the relationship between the symbols in mathematical expressions,a mathematical expression indexing and matching method based on the inter-relevant successive trees was proposed.In order to reduce the growth of key words,in the feature extraction stage,a mathematical expression feature representation method and a clustering method were designed by analyzing the characteristics of LaTeX expression.In the indexing stage,the inter-relevant successive trees index model was applied to the construction of the mathematical expressions index.Through the transformation of the inter-relevant successive tree,the problem of hierarchical growth when the tree structure is used to represent the mathematical expressions could be solved.In the retrieval stage,retrieval algorithms for adaptive index structure were constructed,and a retrieval model was realized which includes accurate retrieval,containment retrieval and fuzzy retrieval.The method was verified by experiments,and the validity of the method was proved.