计算机科学 ›› 2025, Vol. 52 ›› Issue (12): 115-124.doi: 10.11896/jsjkx.241100090
谢文林, 杜明, 周军锋
XIE Wenlin, DU Ming, ZHOU Junfeng
摘要: 枚举概率图上两个顶点间的路径是分析两点间关系的基本手段。为了解决已有算法存在的剪枝不充分、冗余计算等问题,提出了一种基于剪枝和索引的算法——PIEnum,其任务是在概率图G上枚举所有从起点s到终点t,长度不超过k且路径上所有边的概率累积值不低于γ的简单路径,其中k和γ分别为给定的路径长度约束值和概率阈值。对于一个查询,PIEnum首先剔除无效顶点以缩减路径枚举的搜索空间,然后构建一个轻量级的在线索引来避免路径枚举过程中重复的剪枝判断,最后在路径枚举的过程中将无效的搜索分支剪枝。为了进一步提升算法在稠密图上的查询效率,基于Join模式实现了PIEnum+。在10个真实数据集上检验了该算法的性能,实验结果表明,PIEnum整体性能比已有算法提升了10倍以上。
中图分类号:
| [1]LIU H,JIN C,YANG B,et al.Finding top-k shortest paths with diversity[C]//ICDE.IEEE Computer Society,2018:1761-1762. [2]OUYANG D,YUAN L,QIN L,et al.Efficient shortest path index maintenance on dynamic road networks with theoretical guarantees[J].VLDB Endow,2020,13(5):602-615. [3]PENG Y,YU J X,WANG S.PSPC:efficient parallel shortest path counting on large-scale graphs[C]//ICDE.IEEE,2023:896-908. [4]LAI L,QING Z,YANG Z,et al.Distributed subgraph matching on timely dataflow[J].VLDB Endow,2019.12(10):1099-1112. [5]WANG K,LIN X,QIN L,et al.Vertex priority based butterfly counting for large-scale bipartite networks[J].VLDB Endow,2019,12(10):1139-1152. [6]KAPUSTA J,MUNK M,SVEC P.Selection of suitable page-rank calculation for analysis of differences between expected and observed probability of accesses to web page[C]//MIWAI.Springer,2018:139-150. [7]LIBEN D,KLEINBERG J M.The link-prediction problem forsocial networks[J].JASIST,2007,58(7):1019-1031. [8]YAN J,WANG W,WU Z,et al.An uncertain graph method based on shuffle model to preserve link privacy of mobile social networks[J].JISE,2024,40(1):125-150. [9]SEVON P,ERONEN L,HINTSANEN P,et al.Link discovery in graphs derived from biological databases[C]//DILS.Sprin-ger,2006:35-49. [10]KIMURA M,SAITO K.Tractable models for information diffusion in social networks[C]//PKDD.Berlin:Springer,2006:259-271. [11]SAHA A,BROKKELKAMP R,VELAJ Y,et al.Shortest paths and centrality in uncertain networks[J].VLDB Endow,2021,14(7):1188-1201. [12]PENG Y,ZHANG Y,ZHANG W,et al.Efficient probabilistic k-core computation on uncertain graphs[C]//ICDE.IEEE Computer Society,2018:1192-1203. [13]ZHOU A,WANG Y,CHEN L.Butterfly counting and bitruss decomposition on uncertain bipartite graphs[J].VLDB J,2023,32(5):1013-1036. [14]KE X,KHAN A,QUAN L H.An in-depth comparison of s-t reliability algorithms over uncertain graphs[J].VLDB Endow,2019,12(8):864-876. [15]DAI Q,LI R,LIAO M,CHEN H,et al.Fast maximal cliqueenumeration on uncertain graphs:A pivot-based approach[C]//SIGMOD.ACM,2022:2034-2047. [16]YASUDA N,SUGAYA T,MINATO S.Fast compilation of s-t paths on a graph for counting and enumeration[C]//AMBN.PMLR,2017:129-140. [17]BIRMELE E,FERREIRA R A,GROSSI R,et al.Optimal lis-ting of cycles and st-paths in undirected graphs[C]//SODA.SIAM,2013:1884-1896. [18]NISHINO M,YASUDA N,MINATO S,et al.Compiling graph substructures into sentential decision diagrams[C]//AAAI.Springer,2017:1213-1221. [19]GROSSI R,MARINO A,VERSARI L.Efficient algorithms for listing k disjoint st-paths in graphs[C]//LATIN.Springer,2018:544-557. [20]RIZZI R,SACOMOTO G,SAGOT M.Efficiently listing boun-ded length st-paths[C]//IWOCA.Springer,2014:318-329. [21]PENG Y,ZHANG Y,LIN X,et al.Hop-constrained s-t simple path enumeration:Towards bridging theory and practice[J].VLDB Endow,2019,13(4):463-476. [22]QIU X,CEN W,QIAN Z,et al.Realtime constrained cycle detection in large dynamic graphs[J].VLDB Endow,2018,11(12):1876-1888. [23]LAI Z,PENG Y,YANG S,et al.PEFP:efficient k hop constrained s-t simple path enumeration on FPGA[C]//ICDE.IEEE,2021:1320-1331. [24]HAO K,YUAN L,ZHANG W.Distributed hop-constrained s-t simple path enumeration at billion scale[J].VLDB Endow,2021,15(2):169-182. [25]SUN S,CHEN Y,HE B,et al.Pathenum:Towards real-time hop constrained s-t path enumeration[C]//SIGMOD.ACM,2021:1758-1770. [26]ZHANG J,YANG S,OUYANG D,et al.Hop constrained s-t simple path enumeration on large dynamic graphs[C]//ICDE.IEEE,2023:762-775. [27]LI X,HAO K,YANG Z,et al.Hop-constrained s-t simple path enumeration in large uncertain graphs[C]//ADC.Springer,2022:115-127. |
|
||