A Mathematical Indexing Method Based on the Hierarchical Features of Operators in Formulae
Xuedong Tian · 2017
Full text search engines widely used today still have no math searching function, which brings inconvenience for people finding their scientific documents with mathematical query words.It is necessary to research and develop the theory and technology of mathematical expression retrieval.This paper proposed an index model of mathematical expressions for realizing math retrieval through analyzing the characteristics of formulae.Firstly, the FDS data was obtained from the formulae expressed in LaTeX description with recursive analysis.Then, the index features including the level and location features of operators were extracted from the FDS data of formulae.Finally, the extracted features were used to construct a feature vector for dividing formulae into several classes and the math index was constructed for the classes respectively.The experiment was carried out on 134199 formulae and the result shows its effectiveness for improving the efficiency of mathematical expression retrieval.