Research And Implementation Of Metadata Construction And Visualized System For Movable Artifact Based On Semantic Association

The 20th National Congress of the Communist Party of China pointed out the need to "deepen the implementation of the national cultural digitization strategy,promote the innovative development of the integration of excellent traditional Chinese culture and digital technology,and meet the growing spiritual and cultural needs of the people." The digital transformation of museums is a shining wave in the cultural digitalization trend under the strategy of building a socialist cultural regime.This study aims to accelerate the digital transformation of museums by focusing on the collection of cultural relics.Firstly,we constructed a set of movable cultural relic metadata standard system.Secondly,we organized the data of movable cultural relics from multiple sources into a unified and standardized dataset.Then,we adopted multi-label image classification technology to achieve shallow semantic association of the museum’s collection of cultural relics.Finally,we designed and implemented a visualization and retrieval system for the museum’s collection of cultural relics based on force-directed graph layout.The main research contents are as follows:(1)Building a set of standard metadata system for museum collections.Conduct a comparative analysis of typical domestic and foreign metadata standards for cultural relics,and compare the online collection retrieval methods and information content of various museums.Based on this research,construct a set of standard metadata system for museum collections.(2)Building a standardized collection of museum artifact dataset.This study focuses on acquiring museum artifact data through four different channels,cleaning and integrating the data,converting it to a user-friendly format,and constructing a unified and standardized collection of museum artifact dataset.The process includes data cleaning,data integration,and data transformation to store the data in a standardized and easy-to-use format.(3)Implement shallow semantic association of museum collection data based on multi-label classification.Semantic association is the core concept of the cultural gene label semantic model PattemNet,which is used for cultural connotation mining and dissemination.By inputting an image,multiple abstract concepts are generated on the image,representing semantic labels at different levels,and recommending images with similar semantics,which is referred to as semantic association.The research proposes a multi-label image classification algorithm to achieve shallow semantic association,laying the foundation for visualization and retrieval.(4)Building a visualized and searchable system for museum collections.The study explores the calculation of cosine similarity for each image label of the cultural relics,and uses this similarity to perform related searches.At the same time,the cosine similarity of each node is input into the force-directed graph layout algorithm to obtain the visualization of the relationships between nodes,and a visualized and searchable system for museum collections is built.This study attempts to solve the problem of inconsistent data formats and difficult data association between museums,achieve shallow semantic matching,and build a visualized and searchable system for museum collections,which displays the relationship between cultural relics and enables interconnectivity,interoperability and data exchange among museums.