Adaptive Greedy versus Non-adaptive Greedy for Influence Maximization
Authors: Wei Chen, Binghui Peng, Grant Schoenebeck, Biaoshuai Tao
JAIR 2022 | Venue PDF | Archive PDF | Plain Text | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | In this section, we compare the three algorithms the non-adaptive greedy algorithm, the greedy adaptive policy πg and the conservative greedy adaptive policy πg empirically by experiments on the social networks in our daily lives. We implement the experiments on four undirected graphs, shown in Table 2. |
| Researcher Affiliation | Collaboration | Wei Chen EMAIL Microsoft Research, ChinaBinghui Peng EMAIL Department of Computer Science, Columbia UniversityGrant Schoenebeck EMAIL School of Information, University of MichiganBiaoshuai Tao EMAIL School of Electronic Information and Electrical Engineering, Shanghai Jiao Tong University |
| Pseudocode | No | The paper describes algorithms (non-adaptive greedy algorithm, greedy adaptive policy πg, conservative greedy adaptive policy πg) in descriptive text without structured pseudocode or algorithm blocks. |
| Open Source Code | No | The paper does not contain an explicit statement or link indicating the release of source code for the methodology described in this paper. It mentions using or building upon existing reverse-reachable-set-based algorithms but does not provide its own implementation code. |
| Open Datasets | Yes | We implement the experiments on four undirected graphs, shown in Table 2. All of our datasets come from (Leskovec & Krevl, 2014), and these networks are also popular choices in other empirical work. ... Leskovec, J., & Krevl, A. (2014). SNAP Datasets: Stanford large network dataset collection. http://snap.stanford.edu/data. |
| Dataset Splits | No | For each dataset, we sample three realizations φ1, φ2, φ3 as the ground-truth . Therefore, a total of six experiments are performed for each dataset: the two models ICM and LTM for each of the three realizations. This describes different simulation scenarios or realizations for evaluation, not traditional training/test/validation splits of the datasets themselves. |
| Hardware Specification | No | The paper does not provide specific hardware details (e.g., CPU/GPU models, memory) used for running the experiments. It only describes the datasets and models used. |
| Software Dependencies | No | The paper mentions implementing algorithms and refers to other research papers for method details, but does not specify any programming languages, libraries, or software dependencies with version numbers used for the implementation. |
| Experiment Setup | Yes | We implement the three algorithms with k = 200 seeds. For the diffusion model, we implement both ICM and LTM. For ICM, the weight of each edge is set to 0.01. For LTM, each undirected edge (u, v) is viewed as two anti-parallel directed edges such that w(u, v) = 1/ deg(v) and w(v, u) = 1/ deg(u). ... To implement the three algorithms, we sample 1,000,000 reverse reachable sets... For the greedy adaptive policy... 1,000,000 new reverse reachable sets are sampled on the remainder graph. |