This paper analyzes and identifies the composition of ancient glass products by using data analysis method

Abstract

Glass is the precious material evidence of the ancient Silk Road trade. China's ancient glass and foreign glass, although similar in appearance, but the chemical composition is different. Ancient glass is susceptible to environmental influences and thus weathering, and the weathering process will affect the change of its chemical composition, so it is necessary to study the relationship between weathering and some characteristics of glass. First of all, this paper uses the method of Chi-square test to analyze the relationship between the surface weathering of cultural relics and the type, color and decoration of glass, so that there is a greater relationship between the surface weathering and the type. Secondly, the glass is divided into two types: lead barium and high potassium, and the bar chart of each chemical component content of weathered glass and unweathered glass is made. Next, the paper predicts their contents before weathering, and then compares them with the contents of other types and unweathered cultural relics with the same color and pattern to test the prediction results of this paper. Then, this paper uses the K-Means ++ algorithm to classify the unclassified data in Table 2. By analyzing the classification rules obtained by clustering, the classification standards for high-potassium glass and lead-barium glass are approximately obtained. Next, the same clustering algorithm is used to divide the data into high potassium glass group and lead barium glass group, and cluster the two groups of data respectively. Through the analysis of the clustering center law, the classification standard of the subclass is obtained. Finally, the correlation analysis method was used to conduct correlation analysis between the content of main compounds in the subclass classification law of each cultural relic and the distance between each cultural relic and the final clustering center, and the fluctuation of 5% was processed for the content of important compounds, and the correlation analysis was carried out again to test the change of correlation to achieve the purpose of testing sensitivity.