Partitioned Sampling of Public Opinions Based on Their Social Dynamics

Authors: Weiran Huang, Liang Li, Wei Chen

AAAI 2017 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental We use both synthetic and real-world datasets to demonstrate that the partitioned sampling method results in significant improvement in sampling quality and it is robust when some opinion similarities are inaccurate or even missing.
Researcher Affiliation Collaboration Weiran Huang IIIS, Tsinghua University Beijing, China huang.inbox@outlook.com Liang Li AI Department, Ant Financial Group Hangzhou, Zhejiang, China liangli.ll@alipay.com Wei Chen Microsoft Research Beijing, China weic@microsoft.com
Pseudocode Yes Algorithm 1 Greedy Partitioning Algorithm
Open Source Code No The paper does not provide any explicit statement about releasing its source code, nor does it include a link to a code repository.
Open Datasets Yes We use the planted partition model (Condon and Karp 2001) to generate undirected graphs, which aims at resembling the community structure in realworld social networks.
Dataset Splits No The paper describes using synthetic and real-world datasets and mentions "sample size budget r" for sampling, but it does not specify any training, validation, or test dataset splits in terms of percentages, counts, or methodologies for reproducibility.
Hardware Specification No The paper does not provide any specific details about the hardware used to run the experiments, such as CPU or GPU models, or memory specifications.
Software Dependencies Yes We use CVX package (Grant and Boyd 2014; 2008) to solve the SDP programming.
Experiment Setup Yes We describe major parameter settings for the experiments below, while leave the complete settings in the full report (Huang, Li, and Chen 2015) due to space constraints. In our experiment, when the parameters of VIO model are set, the simulation is done by (a) calculating the pairwise opinion similarities by Theorem 4, (b) running the partitioning algorithms to obtain the partition candidate, and (c) computing the expected variance EM[Var S( ˆf)] by Theorem 1.