Notice: The reproducibility variables underlying each score are classified using an automated LLM-based pipeline, validated against a manually labeled dataset. LLM-based classification introduces uncertainty and potential bias; scores should be interpreted as estimates. Full accuracy metrics and methodology are described in [1].
Opinion Maximization in Social Trust Networks
Authors: Pinghua Xu, Wenbin Hu, Jia Wu, Weiwei Liu
IJCAI 2020 | Venue PDF | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | We subsequently formalized two novel problems for solving the issue in STNs. Moreover, we developed two matrix-based methods for these two problems and experiments on real-world datasets to demonstrate the practical utility of our methods. |
| Researcher Affiliation | Academia | 1School of Computer Science, National Engineering Research Center for Multimedia Software and Institute of Artificial Intelligence, Wuhan University 2Shenzhen Research Institute, Wuhan University 3Department of Computing, Macquarie University |
| Pseudocode | Yes | Algorithm 1: Solving internal opinion problem; Algorithm 2: Solving expressed opinion problem |
| Open Source Code | Yes | 1The implementations of the methods are available at https://github.com/WHU-SNA/Op Max In STN |
| Open Datasets | Yes | The real-world social trust networks used in the experiments are the following: (i) Alpha and (ii) OTC [Kumar et al., 2016; Kumar et al., 2018]. We normalized the trust values (i.e., edge weights) to the interval [ 1, 1]. Moreover, we also tested our methods on Elec [Leskovec et al., 2010b; Leskovec et al., 2010a] and Rfa [West et al., 2014], as the relationships in these two networks are closely related to trust. The statistical details of these networks can be found in SNAP.2 (Footnote 2: http://snap.stanford.edu/data/) |
| Dataset Splits | No | The paper describes how internal opinion vectors were initialized and different distributions used for them, but it does not specify explicit training, validation, or test dataset splits (e.g., percentages or counts) for the network data itself. |
| Hardware Specification | Yes | 3On a server with an Intel i9-9820x CPU and 64 GB RAM. |
| Software Dependencies | No | While the paper provides a link to the code implementation, it does not explicitly list the specific software dependencies (e.g., libraries, frameworks) along with their version numbers within the text. |
| Experiment Setup | Yes | To simulate different situations, we randomly sampled values, which obey specific distributions, to initialize the internal opinion vector s. More precisely, for each network, we used five sets of internal opinions. (i) The internal opinions follow a uniform distribution (i.e., s U( 1, 1)). (ii) The internal opinions follow a standard normal distribution (i.e., s N(0, 1)). (iii) The absolute values of the internal opinions follow power-law distributions with α = 1 and α = 2 (i.e., |s| Pow(1) and |s| Pow(2)), and each entry of s is negated with a probability of 0.5. (iv) The internal opinion of a node positively correlates to that node s column connectivity (i.e., si P j |aji|), and each entry of s is negated with a probability of 0.5. |