Distance-Aware DAG Embedding for Proximity Search on Heterogeneous Graphs

Authors: Zemin Liu, Vincent Zheng, Zhou Zhao, Fanwei Zhu, Kevin Chang, Minghui Wu, Jing Ying

AAAI 2018 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental We evaluate D2AGE on three benchmark data sets with six semantic relations, and we show that D2AGE outperforms the state-of-the-art baselines. We release the code on https://github.com/shuai OKshuai.
Researcher Affiliation Academia 1 Zhejiang University, China; 2 Advanced Digital Sciences Center, Singapore; 3 Zhejiang University City College, China; 4 University of Illinois at Urbana-Champaign, USA
Pseudocode Yes Algorithm 1 Distance-aware DAG Generation; Algorithm 2 Distance-aware DAG Embedding
Open Source Code Yes We release the code on https://github.com/shuai OKshuai.
Open Datasets Yes The Linkedin dataset (Li, Wang, and Chang 2014) contains two symmetric relations: schoolmate and colleague; the Facebook dataset (Li, Wang, and Chang 2014) contains two symmetric relations: classmate and family; and the DBLP dataset (Wang et al. 2010) contains two asymmetric relations: advisor and advisee.
Dataset Splits Yes We randomly sample 20% of queries as the training set and 80% as the test set. We repeat this procedure for five times and report the average performance.
Hardware Specification Yes We run our experiments on Linux machines with eight 2.27GHz Intel Xeon(R) CPUs and 32GB memory.
Software Dependencies Yes We use java-1.8 and Theano (Team 2016) for development.
Experiment Setup Yes In the experiments, we set λ = 0.0001, then we tune parameters d, α, β, μ for different datasets.