Fast Solving Maximum Weight Clique Problem in Massive Graphs

Authors: Shaowei Cai, Jinkun Lin

IJCAI 2016 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental Experiments on massive graphs from various applications show that, Fast WClq finds better solutions than state of the art algorithms while the run time is much less.
Researcher Affiliation Academia 1State Key Laboratory of Computer Science, Institute of Software, Chinese Academy of Sciences, Beijing, China 2Key Laboratory of High Confidence Software Technologies, Peking University, Beijing, China
Pseudocode Yes The pseudo code of Fast WClq is shown in Algorithm 1. Algorithm 1: Fast WClq (G, cutoff) ... Algorithm 2: Choose Add V ertex(Cand Set, k)
Open Source Code No The paper states that "Fast WClq is implemented in C++" but does not provide any link or explicit statement about releasing its source code.
Open Datasets Yes The benchmarks in our experiments were originally from the Network Data Repository online [Rossi and Ahmed, 2015],2 including biological networks, collaboration networks, facebook networks, interaction networks, infrastructure networks, amazon recommend networks, retweet networks, scientific computation networks, social networks, technological networks, and web link networks. 2http://www.graphrepository.com/networks.php
Dataset Splits No The paper does not provide explicit details about train/validation/test dataset splits, specific percentages, or sample counts. It refers to using "benchmarks" but not how data within these benchmarks was partitioned for training or validation.
Hardware Specification Yes The experiments are carried out on a workstation under Ubuntu Linux 14.04, using 2 cores of Intel i7-4710MQ CPU @ 2.50 GHz and 32 GByte RAM.
Software Dependencies Yes The experiments are carried out on a workstation under Ubuntu Linux 14.04, using 2 cores of Intel i7-4710MQ CPU @ 2.50 GHz and 32 GByte RAM. All algorithms are complied with g++ version 4.7 with -O3 option.
Experiment Setup Yes Parameters k0 and kmax for dynamic BMS heuristic are set to 4 and 64 (=26). The cutoff time ( ct ) for Fast WClq is 100 seconds per run. For LSCC+BMS, we test it under two cutoff time, 100 and 1000 seconds. For the exact algorithm Max WClq, we run it once on each graph with a cutoff time of one hour.