PolygonE: Modeling N-ary Relational Data as Gyro-Polygons in Hyperbolic Space

Authors: Shiyao Yan, Zequn Zhang, Xian Sun, Guangluan Xu, Shuchao Li, Qing Liu, Nayu Liu, Shensi Wang4308-4317

AAAI 2022 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental Experimental results demonstrate that Polygon E shows SOTA performance on all benchmark datasets and generalizes well on binary data.
Researcher Affiliation Academia 1 Aerospace Information Research Institute, Chinese Academy of Sciences 2 Key Laboratory of Network Information System Technology(NIST), Aerospace Information Research Institute 3 School of Electronic, Electrical and Communication Engineering, University of Chinese Academy of Sciences
Pseudocode No The paper describes the model mathematically but does not include structured pseudocode or an algorithm block.
Open Source Code No The paper does not explicitly state that source code for the described methodology is available, nor does it provide a link to a code repository.
Open Datasets Yes Knowledge base completion (KBC) experiments are conducted on JF17K (Wen et al. 2016), Wiki People (Guan et al. 2019), and FB-AUTO (Fatemi et al. 2020).
Dataset Splits Yes Table 1: Statistics of Datasets (showing #Train, #Valid, #Test columns with specific numbers for each dataset like Wiki People: #Train 305,725, #Valid 38,223, #Test 38,281).
Hardware Specification Yes Experiments are implemented on a single NVIDIA RTX 3080 GPU.
Software Dependencies No The paper does not provide specific software dependencies with version numbers, such as programming language versions or library versions (e.g., PyTorch 1.x).
Experiment Setup Yes Embedding dimensions are set to 50 for a fair comparison with RAM. Other hyper-parameters are chosen from grid search. Concretely, learning rate η is selected from {10, 15, 30, 50, 100}, batch size nbatch are chosen from {64, 128, 256}, number of negative samples nneg are selected from {25, 50, 100}. α and β in equation (13) are integers sampled from {1, 2, 3, 4, 5, 6}. Experiments are implemented on a single NVIDIA RTX 3080 GPU.