Article
Computer Science, Artificial Intelligence
Zezhong Xu, Peng Ye, Juan Li, Huajun Chen, Wen Zhang
Summary: Knowledge reasoning is crucial for overcoming the limitations of knowledge graphs (KGs) and has made significant contributions to the advancement of large KGs. Rule mining, an important task in knowledge reasoning, focuses on learning interpretable inference patterns from KGs. Existing methods mainly concentrate on closed path rules with various relations and variables, disregarding the inclusion of constants. In this paper, we propose EduCe, an Elegant Differentiable rUle learning with Constant mEthod, which considers constants in the rule mining process and incorporates a constant operator and dynamic weight mechanism to enhance rule diversity and accuracy. Experimental results on multiple knowledge graph completion benchmarks demonstrate that EduCe outperforms other differentiable rule mining methods in terms of link prediction and can effectively learn a wide range of high-quality rules.
KNOWLEDGE-BASED SYSTEMS
(2023)
Article
Computer Science, Artificial Intelligence
Penghui Xie, Guangyou Zhou, Jin Liu, Jimmy Xiangji Huang
Summary: FKGC aims to predict missing parts of a query triplet based on a small number of known samples. Existing approaches face challenges in effectively encoding remote neighbor information and modeling uncertainty of few-shot relations. To address these challenges, a global-local neighbor encoding module is proposed, along with an adaptive Gaussian mixture model for modeling few-shot relations.
EXPERT SYSTEMS WITH APPLICATIONS
(2023)
Article
Computer Science, Information Systems
Zepeng Li, Rikui Huang, Minyu Zhai, Zhenwen Zhang, Bin Hu
Summary: This paper proposes a statistic-based algorithm for inferring missing entity types in knowledge graph, considering both performance and incrementality. The algorithm aggregates neighborhood information and type co-occurrence information to infer types, outperforming previous statistics-based algorithms and some other models.
WORLD WIDE WEB-INTERNET AND WEB INFORMATION SYSTEMS
(2023)
Article
Computer Science, Artificial Intelligence
Kamrul Islam, Sabeur Aridhi, Malika Smail-Tabbone
Summary: In this paper, an efficient negative sampling method (SNS) is proposed for learning low dimensional vector representations of KG, improving the performance of link prediction task. Additionally, a rule mining method based on learned embeddings is introduced for analyzing KG and supporting explainable link prediction. Experimental evaluations demonstrate the good performance and potential of these methods.
KNOWLEDGE-BASED SYSTEMS
(2022)
Article
Computer Science, Information Systems
Ling Wang, Jicang Lu, Yepeng Sun
Summary: Existing knowledge graph representation learning (KGRL) models rely on explicit semantic information of triple structure and cannot fully mine the implicit semantic information in the knowledge graph (KG). We propose a novel KGRL model, Melo, that leverages meta-information and logical rules of entities and relations to improve KGRL model performance and accuracy, making up for the disadvantages of existing research. Experimental results demonstrate that Melo enhances performance compared to baselines in terms of multiple evaluation metrics, and visualization methods show how meta-information, logical rules, and triple structure enhance training.
JOURNAL OF KING SAUD UNIVERSITY-COMPUTER AND INFORMATION SCIENCES
(2023)
Article
Computer Science, Artificial Intelligence
Wenqiang Liu, Hongyun Cai, Xu Cheng, Sifa Xie, Yipeng Yu, Dukehyzhang
Summary: The goal of representation learning of knowledge graph is to encode entities and relations into a low-dimensional embedding space. Existing methods have limitations in expressing high-order structural relationships between entities and utilizing attribute triples. To overcome these limitations, this paper proposes a novel method named KANE, which captures high-order structural and attribute information of knowledge graphs using graph convolutional networks. Experimental results show that KANE outperforms other methods.
KNOWLEDGE-BASED SYSTEMS
(2022)
Article
Computer Science, Artificial Intelligence
Bram Steenwinckel, Gilles Vandewiele, Michael Weyns, Terencio Agozzino, Filip De Turck, Femke Ongenae
Summary: This paper presents a novel technique called INK, which learns binary feature-based representations for nodes in a knowledge graph that are comprehensible to humans. By comparing the predictive performances of the node representations obtained through INK with state-of-the-art techniques, such as Graph Convolutional Networks (R-GCN) and RDF2Vec, on both benchmark datasets and a real-world use case, the predictive power of INK is demonstrated.
DATA MINING AND KNOWLEDGE DISCOVERY
(2022)
Article
Computer Science, Artificial Intelligence
Jin Li, Jinpeng Xiang, Jianhua Cheng
Summary: Knowledge graphs are incomplete and knowledge graph completion has become a prominent task to find missing relations. Embedding models simplify operations and enhance knowledge graph completion. We propose a novel method, EARR, which improves the accuracy of knowledge graph completion by separating attributes, using logic rules, and extending the dataset.
EXPERT SYSTEMS WITH APPLICATIONS
(2023)
Article
Chemistry, Multidisciplinary
Yinyu Lan, Shizhu He, Kang Liu, Jun Zhao
Summary: Knowledge graph completion (KGC) is an important research direction for addressing the issue of incomplete knowledge graphs. Traditional rule-based reasoning methods have good accuracy and interpretability, but obtaining effective rules on large-scale knowledge graphs is challenging. Embedding-based reasoning methods have high efficiency and scalability, but they cannot fully utilize domain knowledge in the form of logical rules. This paper proposes a novel method that combines rules and embeddings iteratively to achieve a good balance between efficiency and scalability.
APPLIED SCIENCES-BASEL
(2023)
Article
Computer Science, Artificial Intelligence
Zhen Bi, Siyuan Cheng, Jing Chen, Xiaozhuan Liang, Feiyu Xiong, Ningyu Zhang
Summary: In this paper, we propose a new variant of Transformer called Relphormer for knowledge graph representations. We introduce Triple2Seq to dynamically sample contextualized sub-graph sequences as input, alleviating the heterogeneity issue. We also propose a novel structure-enhanced self-attention mechanism to encode relational information. Experimental results show that Relphormer outperforms baseline models.
Article
Computer Science, Information Systems
Pouya Ghiasnezhad Omran, Kerry Taylor, Sergio Rodriguez Mendez, Armin Haller
Summary: This article introduces a novel algorithm, OPRL, for learning Open Path (OP) rules that can generate relevant queries for Knowledge Graph completion, even when there is no closed rule to answer the query. This demonstrates the first solution for active knowledge graph completion.
INFORMATION SCIENCES
(2022)
Article
Computer Science, Information Systems
Ibrahim A. Ahmed, Fatima N. AL-Aswadi, Khaled M. G. Noaman, Wafa' Za'al Alma'aitah
Summary: With the growth of data on the Web, the need for efficient methods to extract valuable information from the data has increased. Knowledge graphs provide an efficient and easy way to represent and organize data. The construction of Arabic Knowledge Graph (AKG) faces challenges due to limited Arabic data and lack of effective language processing tools. This research reviews KG construction best practices and discusses the challenges and potential solutions in constructing AKG.
JOURNAL OF KING SAUD UNIVERSITY-COMPUTER AND INFORMATION SCIENCES
(2022)
Article
Computer Science, Software Engineering
K. Lakshmi, T. Meyyappan
Summary: This paper discusses the application of complex networks in various scientific disciplines and the challenges of mining important frequent patterns from graph databases. Existing algorithms perform well on medium networks but struggle with large graphs, whereas the proposed algorithm in this paper is efficient and scalable on very large graph databases.
CONCURRENCY AND COMPUTATION-PRACTICE & EXPERIENCE
(2021)
Article
Chemistry, Multidisciplinary
Zhiwei Nie, Shisheng Zheng, Yuanji Liu, Zhefeng Chen, Shunning Li, Kai Lei, Feng Pan
Summary: The paper introduces a semantic representation framework with a dual-attention module for literature mining of LIB cathodes, enabling the detection of deep-seated associations among materials for targeted applications. This work provides a path to text-mining-based knowledge management for complicated materials systems with little dependence on domain expertise.
ADVANCED FUNCTIONAL MATERIALS
(2022)
Article
Computer Science, Artificial Intelligence
Qian Li, Daling Wang, Shi Feng, Kaisong Song, Yifei Zhang, Ge Yu
Summary: This paper proposes the VGAT model and CR protocol to address the prediction of missing links in open knowledge graphs. The VGAT model automatically mines synonymous features using a variational autoencoder densified graph attention mechanism, while the CR protocol comprehensively evaluates multiple answers from the perspectives of significance and compactness.
KNOWLEDGE-BASED SYSTEMS
(2023)