Modeling with Node Degree Preservation Can Accurately Find Communities
Authors: Di Jin, Zheng Chen, Dongxiao He, Weixiong Zhang
AAAI 2015 | Conference PDF | Archive PDF | Plain Text | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | Our results from extensive experiments on synthetic benchmarks and real-world networks show the superior performance of the new method over the existing methods in detecting both disjoint and overlapping communities. To test the performance of our MNDP, we evaluated it on synthetic and real networks. We also compared it with four related methods... |
| Researcher Affiliation | Academia | Di Jin1, Zheng Chen2, Dongxiao He1, Weixiong Zhang2,3 1School of Computer Science and Technology, Tianjin University, Tianjin 300072, China, 2Department of Computer Science and Engineering, Washington University, St. Louis, MO 63130, USA, 3Institute for Systems Biology, Jianghan University, Wuhan 430056, China {jindi, hedongxiao}@tju.edu.cn, {zheng.chen, weixiong.zhang}@wustl.edu |
| Pseudocode | No | The paper describes an algorithm for parameter learning using multiplicative updating rules and mathematical equations (e.g., equation 8), but it does not present this in a structured pseudocode or algorithm block. |
| Open Source Code | No | The paper does not provide any statement or link indicating that the source code for the methodology is openly available. |
| Open Datasets | Yes | To test the performance of our MNDP, we evaluated it on synthetic and real networks...Two types of synthetic benchmarks, one with disjoint communities (Lancichinetti, Fortunato and Radicchi 2008) and the other with overlapping communities (Lancichinetti and Fortunato 2009), were proposed. Here we considered them in our experiments... The real networks analyzed (Newman 2013; Xie, Kelley and Szymanski 2013)... (Newman 2013; Nelson, Mc Evoy and Schreiber 2013). |
| Dataset Splits | No | The paper describes various experimental parameters and evaluation metrics for synthetic and real-world networks, but it does not explicitly provide specific details on training/test/validation dataset splits or cross-validation setups. |
| Hardware Specification | No | The paper does not provide any specific hardware details used for running its experiments. |
| Software Dependencies | No | The paper does not provide specific ancillary software details with version numbers. |
| Experiment Setup | Yes | Following the parameter setting as the LFR benchmark used in (Lancichinetti, Fortunato and Radicchi 2008), we considered networks with 1000 nodes and the minimum community size cmin of 10 or 20. We varied the mixing parameter μ, which specifies the fraction of the links of a node connecting to nodes outside of the node s community, from 0 to 0.8 with an increment of 0.05. The remaining parameters were kept fixed: the average degree d was set to 20, the maximum degree dmax to 2.5 d, the maximum community size cmax to 5 cmin, the exponent of power-law distribution of node degrees τ1 to -2 and community size τ2 to -1. (Section 4.1.1) and In our experiments, we first get an initial value of by setting ߣ= 0. Then we restart the optimization with = and let ߣ to a relatively large number, e.g., 1000, to minimize the chance of violating the degree constraints. (Section 3.3) |