Rapid Clustering Method Of Large-scale Internet Geographic Markers

As Internet technology continues to mature and spread, internet GIS services hasmade rapid development, the geographic markers based internet electronic map has been widelyused, with a large number of users, internet geographic markers has a huge amount of data, andupdated frequently. Face the mass of data, people obtains information, processes informationthrough its own, however, there are much potential knowledge and law hidden behind the largeamounts of data and they are need for people to discover, to dig, and people urgently need toknow is precisely the knowledge and the law rather than the data itself.Spatial clustering technology is emerged as aninstrument in response to spatial data mining.With the3S technology continues to evolve and mature, a large number of spatial data withcomplex properties is continuously to be collected, and the internet geographic markers is oneof them. Through the use of spatial clustering technology to analysis a large number of data isan extremely effective means to discover the information and the knowledge contained behindthe massive amounts of data, the spatial clustering of massive geographic markers caneffectively find the spatial distribution,accumulation features, spatial trends and spatialassociation of spatial objects.In this paper, against the height heterogeneity of the internet geographic markers, designs aunified management and storage model, and taking into account the high massdata storage, fast concurrent access and query needs,using non-traditional relationship-based, scalable document database MongoDB to store massiveamounts of markers’ data, at the same time, by analysising the capacity of the traditional spatialclustering methods of processing massive data, uses a fast mass data clustering method based onmulti-level clustering unit, to achieve mapping the vast amounts of data tomultiple cluster units, calculations the clusters by clustering units as thecomputing units, thereby can greatly reducing the computation andimprove operational efficiency.