Minimizing Polarization and Disagreement in Social Networks via Link Recommendation

Authors: Liwang Zhu, Qi Bao, Zhongzhi Zhang

NeurIPS 2021 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental Extensive experiments on real datasets demonstrate both the efficiency and effectiveness of our algorithms.
Researcher Affiliation Academia Liwang Zhu, Qi Bao, and Zhongzhi Zhang Shanghai Key Lab of Intelligent Information Processing, Fudan University, Shanghai, China School of Computer Science, Fudan University, Shanghai 200433, China {19210240147, 20110240002, zhangzz}@fudan.edu.cn
Pseudocode Yes Algorithm 1: SPGREEDY(G, EC, k, s); Algorithm 2: COMP(G, EC, s, ϵ); Algorithm 3: FASTGREEDY(G, EC, s, k, ϵ)
Open Source Code No The paper states: 'Omitted proofs and implementation are provided as supplementary material.' but does not provide a direct link or explicit confirmation of code availability for the main algorithms (SPGREEDY, FASTGREEDY). It only provides a link to a third-party solver used: 'the Julia implementation of which is available at https://github. com/danspielman/Laplacians.jl.'
Open Datasets Yes The studied realistic networks are representatively selected from various domains, which are publicly available in the KONECT [31] and SNAP [33].
Dataset Splits No The paper does not provide specific training/validation/test splits, as the problem is formulated as an optimization task on existing networks rather than a traditional machine learning classification or regression task.
Hardware Specification Yes All experiments were conducted on a machine equipped with 32G RAM and 4.2 GHz Intel i7-7700 CPU.
Software Dependencies No The paper states, 'All algorithms in our experiments are executed in Julia.' and mentions 'the SDDM solver SOLVE [44], the Julia implementation of which is available at https://github. com/danspielman/Laplacians.jl.' However, specific version numbers for Julia or the Laplacians.jl package are not provided in the text.
Experiment Setup Yes For the approximation algorithm FASTGREEDY, we set ϵ = 0.3. ... We investigate three different distributions for the initial opinions: uniform distribution, exponential distribution, and power-law distribution.