Quantitative Mining and Genealogy Construction of Aesthetic Features of Ancient Chinese Porcelain Patterns Driven by Big Data

Porcelain, as a precious heritage in the treasure house of Chinese civilization, has become a key technology for promoting its protection and inheritance under the trend of digitalization and intelligence of cultural heritage. To improve the matching accuracy between the visual aesthetic features of ancient Chinese porcelain patterns and the corresponding textual information, this paper proposes a cross-modal entity alignment strategy based on pattern aesthetics within the visual-language pre-training framework. On this basis, a porcelain cultural knowledge graph platform based on big data is designed and developed, which integrates multiple core functional modules such as intelligent interactive questioning, speech synthesis, pattern aesthetic text generation, entity recognition, and relationship extraction. The actual system operation test shows that the average recognition accuracy exceeds 90% and the average response time is less than 5 seconds, verifying the efficiency and reliability of the system in practical applications.