Latent Mapping Semi-Supervised Relationship Community Atmospheric Pollutants Recycling Discovery Based on Ecological Protection and Pagerank Algorithm

Ecological construction and environmental protection are the ecological environment reproduction activities carried out by human beings engaged in living labor and materialized labor. A multi-angle identification controllable community discovery algorithm (PRSGMFCA) based on improved PageRank algorithm and Spin-glass model is proposed aiming at how to effectively solve the overlap problem in community discovery algorithm and how to improve the community expansion model. Firstly, each node is ranked by the optimized PageRank algorithm, and the core node is determined. Secondly, the expansion mode of local community is optimized by multi-angle recognition modularity based on SPIN-glass model, which solves the restriction and influence of traditional modularity on resolution limit. Secondly, the breakpoint design and sequential search are carried out for the optimal value partition of parameters, and the range of parameters is further limited. All parameters correspond to unique reduction weights, and parallelize multi-partition synchronization parameter optimization, further facilitating the equalization and integration of multiple partition information; finally, the validity of the algorithm in relational user community detection is verified by experimental comparison based on Sina's relational community discovery data.