Application of Graph Data Mining Technology in the Construction of Marine Talent Team

Abstract:
Construction marine talent team is still under research. This paper proposes a method to quickly build a marine talent team by mining the marine talent graph data constructed by the author's cooperative relationship. We construct a weighted talent relationship network graph with attribute information, and design an attribute-constrained dense subgraph mining algorithm CP-Qclique based on color prunng strategy. According to the input specific attributes, the algorithm can quickly dig out the talent team suitable for completing specific tasks from the talent relationship network graph. We experiment by crawling DBLP's nearly ten-year dataset. The experimental results show that our proposed CP-Qclique team query algorithm is faster than the basic query algorithm BA-Qclique.