Abstract: A cross-media knowledge semantic representation method and apparatus. The method comprises: performing data acquisition according to a preset semantic description; inputting data information of a topological structure acquired by the data acquisition into a preset stack of an automat corresponding to the semantic description, the finite state set is used for indicating states included in the automat, and the input vocabulary list is used for indicating vocabularies included in the automat; mapping the data information by the automat to obtain key frames corresponding respectively to substructures and/or branches of a target object acquired by the data acquisition; and generating a visual semantic representation of the topological structure according to the key frames corresponding respectively to the substructures and/or branches of the target object acquired by the data acquisition, such that cross-media knowledge alignment is realized.

Abstract: A method and an apparatus for cross-media corresponding knowledge generation.