Guidelines for Experimental Algorithmics: A Case Study in Network Analysis

Angriman, Eugenio; Grinten, Alexander van der; Looz, Moritz von; Meyerhenke, Henning; Nöllenburg, Martin; Predari, Maria; Tzovas, Charilaos

doi:10.3390/a12070127

Cited by 18 publications

(19 citation statements)

References 96 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…For each graph we repeat the experiment three times and we set the imbalance tolerance to 3%, one of the values used in [36] and in ParMETIS. To ensure reproducibility, all experiments were managed by SimexPal [2]. Our code and the experimental pipeline can be found at https://github.com/hu-macsy/KaHIP.…”

Section: Methodsmentioning

confidence: 99%

An MPI-based Algorithm for Mapping Complex Networks onto Hierarchical Architectures

Predari¹,

Tzovas²,

Schulz³

et al. 2021

Preprint

Self Cite

View full text Add to dashboard Cite

Processing massive application graphs on distributed memory systems requires to map the graphs onto the system's processing elements (PEs). This task becomes all the more important when PEs have non-uniform communication costs or the input is highly irregular. Typically, mapping is addressed using partitioning, in a two-step approach or an integrated one. Parallel partitioning tools do exist; yet, corresponding mapping algorithms or their public implementations all have major sequential parts or other severe scaling limitations. In this paper, we propose a parallel algorithm that maps graphs onto the PEs of a hierarchical system. Our solution integrates partitioning and mapping; it models the system hierarchy in a concise way as an implicit labeled tree. The vertices of the application graph are labeled as well, and these vertex labels induce the mapping. The mapping optimization follows the basic idea of parallel label propagation, but we tailor the gain computations of label changes to quickly account for the induced communication costs. Our MPI-based code is the first public implementation of a parallel graph mapping algorithm; to this end, we extend the partitioning library ParHIP. To evaluate our algorithm's implementation, we perform comparative experiments with complex networks in the million-and billion-scale range. In general our mapping tool shows good scalability on up to a few thousand PEs. Compared to other MPI-based competitors, our algorithm achieves the best speed to quality trade-off and our quality results are even better than non-parallel mapping tools.

show abstract

Section: Methodsmentioning

confidence: 99%

An MPI-based Algorithm for Mapping Complex Networks onto Hierarchical Architectures

Predari¹,

Tzovas²,

Schulz³

et al. 2021

Preprint

Self Cite

View full text Add to dashboard Cite

show abstract

“…Our algorithms are implemented in the NetworKit [20] C++ framework and use PCG32 [21] to generate random numbers. All experiments were managed by the SimexPal software to ensure reproducibility [22]. Experiments were executed with sequential code on a Linux machine with an Intel Xeon Gold 6154 CPU and 1.5 TiB of memory.…”

Section: Methodsmentioning

confidence: 99%

Local Search for Group Closeness Maximization on Big Graphs

Angriman

Grinten

Meyerhenke

2019

2019 IEEE International Conference on Big Data (Big Data)

Self Cite

View full text Add to dashboard Cite

In network analysis and graph mining, closeness centrality is a popular measure to infer the importance of a vertex. Computing closeness efficiently for individual vertices received considerable attention. The N P-hard problem of group closeness maximization, in turn, is more challenging: the objective is to find a vertex group that is central as a whole and state-ofthe-art heuristics for it do not scale to very big graphs yet.In this paper, we present new local search heuristics for group closeness maximization. By using randomized approximation techniques and dynamic data structures, our algorithms are often able to perform locally optimal decisions efficiently. The final result is a group with high (but not optimal) closeness centrality.We compare our algorithms to the current state-of-the-art greedy heuristic both on weighted and on unweighted real-world graphs. For graphs with hundreds of millions of edges, our local search algorithms take only around ten minutes, while greedy requires more than ten hours. Overall, our new algorithms are between one and two orders of magnitude faster, depending on the desired group size and solution quality. For example, on weighted graphs and k = 10, our algorithms yield solutions of 12.4% higher quality, while also being 793.6× faster. For unweighted graphs and k = 10, we achieve solutions within 99.4% of the state-of-the-art quality while being 127.8× faster.1 While this greedy algorithm was claimed to have a bounded approximation quality (e. g., in [7], [12]), the proof of this bound relied on the assumption that C is submodular. A recent update [13] to the conference version [12] revealed that, in fact, C is not submodular. We are not aware of any approximation algorithm for group closeness that scales to large graphs.

show abstract

“…Graph processing suffers from methodological issues similar to other computing disciplines. 5,24 Running comprehensive graph processing experiments, especially at scale, lacks tractability 9 -that is, the ability to implement, deploy, and experiment within a reasonable amount of time and cost. As in other computing disciplines, 5,24 we need new, reproducible, experimental methodologies.…”

Section: Ecosystemsmentioning

confidence: 99%

“…We envision a combination of approaches. As in other computing disciplines, 5,24 we need new, reproducible experimental methodologies. Concrete questions arise: How do we facilitate quick yet meaningful performance testing?…”

Section: Resource Managersmentioning

confidence: 99%