[ Purpose/significance] This paper aims to narrow the semantic gap in the museum images retrieval system and explore the value of applying the social tags categorization mechanism to museum, to promote the virtual display of cultural heritage in the museum and enhance the access rate. [ Method/process ] The categorization framework of images retrieval expression is expanded, the categorization model of social tags is created, the classification system of social tags is developed, and the data is analyzed. [ Result/conclusion] The results shows that users prefer to describe the subject of image...