Community detection involves grouping nodes in a graph with dense connections within groups, than between them. We previously proposed efficient multicore (GVE-LPA) and GPU-based ($\nu$-LPA) implementations of Label Propagation Algorithm (LPA) for community detection. However, these methods incur high memory overhead due to their per-thread/per-vertex hashtables. This makes it challenging to process large graphs on shared memory systems. In this report, we introduce memory-efficient GPU-based LPA implementations, using weighted Boyer-Moore (BM) and Misra-Gries (MG) sketches. Our new implementation, $\nu$MG8-LPA, using an 8-slot MG sketch, reduces memory usage by 98x and 44x compared to GVE-LPA and $\nu$-LPA, respectively. It is also 2.4x faster than GVE-LPA and only 1.1x slower than $\nu$-LPA, with minimal quality loss (4.7%/2.9% drop compared to GVE-LPA/$\nu$-LPA).

翻译：社区发现旨在将图中的节点划分为若干组，使得组内连接稠密而组间连接稀疏。我们先前提出了用于社区发现的高效多核（GVE-LPA）与基于GPU（$\nu$-LPA）的标签传播算法（LPA）实现。然而，这些方法因采用每线程/每顶点的哈希表而产生较高的内存开销，使得在共享内存系统上处理大规模图面临挑战。本报告介绍了基于GPU的内存高效LPA实现，其采用了加权Boyer-Moore（BM）与Misra-Gries（MG）草图。我们提出的新实现$\nu$MG8-LPA使用8槽MG草图，与GVE-LPA和$\nu$-LPA相比，分别将内存使用量降低了98倍和44倍。同时，其速度比GVE-LPA快2.4倍，仅比$\nu$-LPA慢1.1倍，且质量损失极小（与GVE-LPA/$\nu$-LPA相比，性能下降分别为4.7%/2.9%）。