Research On The Link Prediction Algorithm Based On Community Comprehensive Property

With the rapid development of the Internet as the representative of network information technology, human society has entered the era of complex networks. Human life and production activities are increasingly dependent on complex systems. Science is both a natural extension of the classical graph theory and random graph theory in mathematics, but also a innovation and development of system science and complexity science. Scholars researched complex network involved in more and more disciplines such as physics, biology, economics, etc., by the topology between network nodes to discover the unknown or the future will be connected to this issue has become increasingly important research point.Link prediction of complex network refers to the unknown links or the forecast the future links.This paper describes the inevitability and importance of community in real networks exist during link prediction, starting from the node where the community, when the node where the community is getting closer, the possibility that exists between them and the more even edges be greater. And even among the community side of connectivity between the community it has a crucial role, and the greater the amount of similarity in communication between the greater community. Therefore, in the community when the similarity calculation, based on the community modularity, combined with the importance of inter-community even edges. In the process of resource allocation, the status of the nodes in the community resource allocation node, that similarity also plays an important role. Finally, a combination of similarity similarity node itself, and the node where the community link prediction algorithm integrated community property is determined based on a comprehensive community property. In order to verify the effectiveness of the algorithm, the algorithm is applied to five real network, obtain network topology analysis features, and Matlab simulation experiments comparing the accuracy of the algorithm and the traditional algorithm AUC, and compared the sensitivity of the algorithm and the actual operation time with other algorithms, on the whole better than others, you can use the integrated community property prediction algorithms based on the link in the future related research.