郑州—洛阳地区史前连续文化聚落的K-means聚类挖掘研究

Research on K-means Clustering Mining for Prehistoric Continuous Culture Settlements about Zhengzhou-Luoyang Region

利用K-means聚类算法和GIS组件SuperMap Objects,开发了基于连续文化序列的空间数据挖掘系统,运用该系统对郑州—洛阳地区史前3个连续文化时期的聚落进行数据挖掘,提取聚落遗址群的聚类规则。通过对聚类规则的分析可知,在仰韶文化后期和龙山文化时期,那些靠近聚落群中心且面积较大的高等级聚落,在位置和交通方面都具有很大的优越性,具备中心聚落的特点,成为中心聚落的潜力大,有发展成为王都的趋势。

Using K-means clustering algorithm and GIS components Super Map Objects, the spatial data mining system based on continuous culture sequence is developed. The system is applied to the data mining for the prehistoric settlements during the three culture phases of Zhengzhou-Luoyang region; the clustering rules of the sites are excavated. It is showed that the high level settlements which are nearby the cluster center in the last period of Yangshao culture and Longshan culture phase, have a lot of advantages in location and traffic, and have characteristics of central settlement by analyzing the clustering rules. These large area settlements have great potential to become a city.