Individual-centered Partial Information in Social Networks
Authors: Xiao Han, Y. X. Rachel Wang, Qing Yang, Xin Tong
JMLR 2024 | Venue PDF | Archive PDF | Plain Text | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | Using simulated and real networks, we demonstrate the performance of our algorithms and compare our centrality measure with other popular alternatives to show it captures unique nodal information. |
| Researcher Affiliation | Academia | International Institute of Finance, School of Management University of Science and Technology of China Hefei, 230026, China School of Mathematics and Statistics University of Sydney NSW, 2006, Australia Department of Data Sciences and Operations, Marshall School of Business University of Southern California CA, 90089, USA |
| Pseudocode | Yes | We summarize this procedure in Algorithm 1. Algorithm 2 Community detection under the SBM Algorithm 3 Community detection under the DCSBM Algorithm 4 Community detection under the SBM (faster alternative for matching) |
| Open Source Code | No | No explicit statement about code release or repository link found in the paper. |
| Open Datasets | Yes | take the well-known Zakary’s karate club data. This dataset contains information about the social interactions and the diffusion of information about a microfinance program in 43 Indian villages (Banerjee et al., 2013; Cheng et al., 2021). The political blog network (Adamic and Glance, 2005) records hyperlinks between web blogs observed in the run-up to the 2004 U.S. presidential election. |
| Dataset Splits | No | For every combination of n and q, we simulate 100 datasets; Algorithm 2 is applied to the partial network centered at node 1 in each dataset. We apply Algorithm 3 to each household in all the villages. we pick six individuals (blogs) and examine their network information and clustering performance in detail. No explicit mention of training/test/validation splits for any dataset is provided. |
| Hardware Specification | No | No specific hardware details (e.g., GPU/CPU models, memory) are mentioned for running experiments. |
| Software Dependencies | No | No specific software dependencies with version numbers are mentioned in the paper. |
| Experiment Setup | Yes | For Model 1, we vary the number of individuals n {300, 600, 900, 1200, 1500, 1800, 2100} and the edge density q {.1, log n/n, (log n/n)1/4/2, 1/ n}. Apply the k-means algorithm to the rows {W(i) : i [n], a1i = 1} and {W(i) : i [n], a1i = 0}, respectively, to separate each group into K clusters. Apply the spherical k-median algorithm (22) to the non-zero rows of SW and (I S)W respectively. |