Attribute Induction-Oriented Excavation and Generalization Analysis of Site Archaeological Data

Abstract

A significant amount of digital archaeological data has emerged as a result of the recent increase in archaeological activity, which is crucial for the preservation of cultural heritage. However, redundant and repetitive archaeological data information often leads to difficulties in management. The study first enhances the Apriori algorithm, which is based on determining the artefact attributes of site archaeological data by applying the boosting degree with difference, in order to increase the effectiveness of archaeological research. A K-means algorithm with adaptive selection of initial clustering centres was then proposed as a means to generalise the archaeological data for analysis. The outcomes revealed that the enhanced Apriori algorithm's maximum runtime was only 0.33 seconds and its minimum runtime was 0.1 seconds. Due to the low impact of noise points on the dataset Flame, the revised K-means algorithm's standard deviation is only 2.537, with the majority of the error values being clustered around zero. After combining the two methods, the classification accuracy of the digitised resources of the site is concentrated around 92%, with high classification accuracy and data generalisation processing ability, which improves the processing efficiency and provides a more reliable method reference for site archaeological research efficiency improvement.