4 results about "Community mining" patented technology

This invention discloses a method and apparatus for assessing the importance of stations, belonging to the field of transportation networks and propagation dynamics technology. It addresses the problem that existing methods for assessing station importance fail to fully reflect operating routes and rely on global network topology information based on shortest paths, leading to relatively time-consuming calculations of the impact of distant stations on station importance. The method includes: constructing a bus-metro dual-layer network, a node network, a route network, and a bipartite graph network based on the bus and metro networks; determining the community structure and community center nodes of the node network using a community mining algorithm to form an enhanced node network; determining the preliminary importance of each second node based on the enhanced node network, and determining the preliminary importance of each third node based on the route network; and determining the final importance of the second nodes included in the node network based on the bipartite graph network, the preliminary importance of each second node, and the preliminary importance of each third node.

The application discloses a hierarchical spatial community mining method based on Gosper-island and belongs to the field of spatial data mining. First, the application carries out hierarchical partitioning of space by adopting Gosper-island, uses hierarchical partitioning coding to record the spatial hierarchical constraint of a partitioning unit, and uses Gosper space filling coding to record the spatial proximity constraint of the same hierarchical partitioning unit. Then, the hierarchical spatial network is constructed by taking the partitioning unit as a network node, the network is organized in the form of a network adjacency matrix, the network adjacency matrix heat map is drawn according to the Gosper coding sequence of the node, and finally, the hierarchical matrix block structure of the network heat map at different levels is extracted from top to bottom to obtain the hierarchical spatial community structure meeting the spatial hierarchical constraint. The application conforms to the hierarchical characteristics of the hierarchical spatial community and can effectively avoid the problems caused by the partitioning hierarchical tree.

The application provides a shared bicycle travel network community mining method based on deep learning. The method comprises the following steps: analyzing and counting the road network data information, the shared bicycle data information and the travel data of users in a specified area, and constructing a traffic travel network; quantitatively describing the travel characteristics of users in the shared bicycle travel network, taking the user travel quantitative indicators as network measurement indicators of a graph neural network model, and constructing the graph neural network model; performing community clustering analysis based on the graph neural network model by using a ClusterNet algorithm, and mining community structures. According to the method, a dynamic travel network is constructed according to different time periods, and the spatiotemporal distribution of travel demands in different stages is analyzed. By using a spatial statistics and a complex network method, indicators are constructed to quantitatively describe the travel characteristics, so that the changes of the travel modes of users in different periods can be clearly understood, and the community structures in the travel network can be dynamically mined.

The invention relates to the technical field of complex network analysis and graph machine learning, in particular to a community detection method based on theme emotion and a weak edge optimization strategy. The method aims at solving the problems that an existing community detection technology is sensitive to weak connection edges and noise data in social network application, multi-dimensional joint optimization such as theme consistency cannot be achieved, and implicit semantic relations are difficult to capture. A technical framework comprising a data preprocessing module, a theme-emotion enhancement module, a weak edge optimization module and a multi-target community division module is constructed; through multi-modal data fusion, topic emotion collaborative modeling, weak edge dynamic optimization and multi-objective function optimization, double improvement of community structure compactness and semantic consistency is realized, noise interference is effectively suppressed, the accuracy, robustness and interpretability of community detection are remarkably improved, and the method is suitable for hidden community mining of a large-scale social network.