A Retrieval Model Of Mathematical Expressions Based On Hierarchical Structures Of Formulae

Mathematical expression is an important description form of scientific and technological information. Realizing the retrieval of mathematical expressions is helpful to exchange and utilize scientific and technical information. However, the complex two-dimensional structure of mathematical expressions results in the inefficiency of retrieving mathematical expressions through tradition full-text retrieval technology. Focused on this problem, a hierarchical index model as well as retrieval methods were proposed based on the hierarchical structure of mathematical expressions. During the research, formulae's baseline structures were extracted based on which the key level was defined. Under the situation, formulae could be clustered in accordance with the characteristics of their key levels. Then, based on the clusters, the data structure of Treap was applied to handle the mathematical expressions resources, meanwhile the model of Inverted Files utilized to organize the files information. At the same time, the hierarchical index structure focusing on formulae matching and searching files information had been formed. In addition, the retrieval algorithms include accurate retrieval algorithm and structural retrieval algorithm, which were adapted to the hierarchical index model. Experiments on the hierarchical index model and retrieval algorithms suggested that the proposed index model and algorithms were efficiency and adaptability in handling the problems caused by two-dimensional structure of mathematical expressions.