A Gravitation-based Position Model for Visualizing and Clustering Vectored Resources

How to retrieve web resources efficiently is an important issue in the web data deluge age. Information visualization and clustering are important techniques for large scaled information organization and retrieval, and vectors have been used for semantic representation of resources. This paper proposes a novel approach to positioning, visualizing and clustering web resources by a gravitation based model, called semantic filed model, which is a disc with several gravitation sources in the circumstance to locate a resource by calculating its position in the disc. Experiments on clustering resources show that the gravitation based position model is efficient and effective for information visualization and resource clustering.