APKGC: Noise-enhanced Multi-Modal Knowledge Graph Completion with Attention Penalty
Authors: Yue Jian, Xiangyu Luo, Zhifei Li, Miao Zhang, Yan Zhang, Kui Xiao, Xiaoju Hou
AAAI 2025 | Venue PDF | Archive PDF | Plain Text | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | Experimental evaluation demonstrates that APKGC excels in overcoming these challenges. Compared to the existing state-of-the-art MMKGC model, APKGC improves Hit@1 by 3.3% on the DB15K dataset and by 3.4% on the MKG-W dataset. We perform extensive experiments on two separate public datasets. |
| Researcher Affiliation | Academia | 1School of Computer Science, Hubei University, Wuhan 430062, China 2Hubei Key Laboratory of Big Data Intelligent Analysis and Application (Hubei University), Wuhan 430062, China 3Key Laboratory of Intelligent Sensing System and Security (Hubei University), Ministry of Education, Wuhan 430062, China 4School of Cyber Science and Technology, Hubei University, Wuhan 430062, China 5Institute of Vocational Education, Guangdong Industry Polytechnic University, Guangzhou 510300, China EMAIL, EMAIL |
| Pseudocode | No | The paper describes the methodology using text and equations (e.g., equations 1-20) and diagrams (Figure 2), but does not include any explicitly labeled pseudocode or algorithm blocks. |
| Open Source Code | Yes | Code https://github.com/Hubu KG/APKGC |
| Open Datasets | Yes | In this study, we utilize two publicly available MMKGC benchmarks, DB15K (Liu et al. 2019) and MKGW (Xu et al. 2022a), to evaluate model performance. |
| Dataset Splits | Yes | Table 1: Statistics of the experimental datasets. # Train 79,222 34,196 # Valid 9,902 4,276 # Test 9,904 4,274 |
| Hardware Specification | Yes | The APKGC model utilizes Py Torch [47] and operates on an Nvidia RTX 4090 GPU. |
| Software Dependencies | No | The APKGC model utilizes Py Torch [47] and operates on an Nvidia RTX 4090 GPU. While PyTorch is mentioned, a specific version number is not provided, nor are any other software dependencies with version numbers. |
| Experiment Setup | Yes | Parameter Settings: The APKGC model utilizes Py Torch [47] and operates on an Nvidia RTX 4090 GPU. For the training, the embedding dimension is set to 128, the batch size is set to 2048, the number of negative samples is set to 64, and the learning rate is set to 1e-4. For the DB15K dataset, Gaussian noise sampling is employed, while for the MKG-W dataset, average noise sampling is utilized. |