A knowledge graph completion model based on contrastive learning and relation enhancement method

Li, LinYu; Zhang, Xuan; Ma, Yubin; Gao, Chen; Wang, Jishu; Yu, Yong; Yuan, Zihao; Ma, QiuYing

doi:10.1016/j.knosys.2022.109889

Cited by 21 publications

(4 citation statements)

References 34 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Subsequently, representation learning models based on GNNs and GATs have been proposed. For instance, R-GCN [31] and ComplexGCN [32] handle node relations based on convolution, while KBGAT [33] and MRGAT [34] respectively consider attention mechanisms and heterogeneous multiple relation connections when aggregating neighboring information. While deep learning-based models demonstrate excellent predictive capabilities between unknown entities and relations, they come with high computational costs and training difficulties.…”

Section: Deep Learning-based Modelsmentioning

confidence: 99%

“…These models include ATTH [38], HyperGEL [39], MuRP [36], and Hybonet [42]. Additionally, we will compare them with state-of-the-art and representative Euclidean and complex space embedding models that have proposed knowledge graph reasoning methods, including TransE [21], DistMult [26], MuRE [36], TuckER [28], ConvE [29], ConvKB [30], and KBGAT [33] for Euclidean space, and ComplEx [27], RotatE [23], and ComplexGCN [32] for complex space. Furthermore, we will consider graph neural network based models and recent models with outstanding predictive performance, such as R-GCN [31], MRGAT [34], and GTKGC [43].…”

Section: Baselinesmentioning

confidence: 99%

See 1 more Smart Citation

Knowledge Graph Link Prediction via Hyperbolic Attenuated Attention Networks

Zhang,

Wu,

Peng

et al. 2024

Preprint

View full text Add to dashboard Cite

Graph Attention Networks serve as prevalent models for knowledge graph link prediction, enabling the completion of link prediction tasks on general knowledge graphs. However, most current models embed nodes in Euclidean space, leading to distortion in node embedding features. Additionally, existing methods overlook the issue of assigning weights to n-hop neighbors during node embedding aggregation and fail to investigate the prediction of different relation patterns in hyperbolic space. In this paper, we propose a Hyperbolic Graph Attenuated Attenuation Network (HGAAT) model for link prediction, utilizing hyperbolic graph attention networks to handle relations and entity representations in knowledge graphs. By introducing a attenuated attention mechanism, HGAAT effectively integrates information from n-top neighboring nodes, thereby generating more accurate entity and relation embeddings in complex relational graphs. Extensive experiments demonstrate the effectiveness and superiority of the HGAAT model in knowledge graph link prediction tasks.

show abstract

Section: Deep Learning-based Modelsmentioning

confidence: 99%

Section: Baselinesmentioning

confidence: 99%

Knowledge Graph Link Prediction via Hyperbolic Attenuated Attention Networks

Zhang,

Wu,

Peng

et al. 2024

Preprint

View full text Add to dashboard Cite

show abstract

“…Moreover, we did not incorporate multi-features such as phonetics information (another type of sequence information) [41] or graph technology to extract latent entity relationships containing auxiliary knowledge [45]. Moreover, a foreseeable challenge in integrating multifeatures into Word-unit BLS is the substantial increase in training time.…”

Section: B Limitationsmentioning

confidence: 99%

Enhanced Word-Unit Broad Learning System With Sememes

Jiang,

Lu,

et al. 2023

IEEE Access

View full text Add to dashboard Cite

High accuracy in text classification can be achieved by simultaneously learning multiple sources of information, such as sequence and word. In this study, we propose a novel learning framework for text classification, called Word-unit Broad Learning System (BLS). The Word-unit BLS utilises a flat neural network known as BLS and offers three key advantages. First, it provides higher accuracy and shorter training time compared to popular machine learning methods, allowing for the simultaneous learning of sequence information and word importance. Second, we incorporate a multi-layer perceptron with attention aggregation in the feature-mapped layer, along with position encoding, to capture the latent relationship between each word and the contextual information in a global context. Lastly, we introduce a novel approach to enhance word representation by employing sememes in the enhancement node layer, thereby improving the feature distribution of each word in the vector space. The effectiveness of the proposed framework was evaluated by conducting experiments on four datasets covering various types of text classifications. The results demonstrate that Word-unit BLS achieves 8.26% higher accuracy than Naive Bayes while requiring 1 33 ⁄ of the training time. Furthermore, when compared with traditional BLS models, Word-unit BLS outperforms in learning the sequence information. The effectiveness of sememe enhancement in word representation is also demonstrated, particularly in the case of large-scale datasets.

show abstract

“…For translation models and bilinear models, the models are relatively simple, and it is difficult to fully explore the relationship between triple entities. Neural network models and rotation models often need to spend more memory space to express the embeddings of entities to obtain more semantic information [12][13][14]. In order to reduce the demand for memory space, in recent years, hyperbolic models have gradually attracted attention in hyperbolic space [15,16].…”

Section: Introductionmentioning

confidence: 99%

Knowledge Graph Completion Based on Entity Descriptions in Hyperbolic Space

2022

View full text Add to dashboard Cite

Hyperbolic space has received extensive attention because it can accurately and concisely represent hierarchical data. Currently, for knowledge graph completion tasks, the introduction of exogenous information of entities can enrich the knowledge representation of entities, but there is a problem that entities have different levels under different relations, and the embeddings of different entities in Euclidean space often requires high dimensional space to distinguish. Therefore, in order to solve the above problem, we propose a method that use entity descriptions to complete the knowledge graph in the Poincaré ball model of hyperbolic space. In this method, the text representation of entity descriptions is in Euclidean space and mapped to hyperbolic space through exponential map. Next, the triple embeddings are initialized in hyperbolic space, and the structured representation of the triple is trained by the hyperbolic model. Finally, the text representation and the structured representation of the entity are cross-fused in hyperbolic space, and then the balance factors are used to adjust the unbalanced energy function. Experimental results show that, compared with baseline models, the proposed method can improve the performance of knowledge graphs completion.

show abstract

A knowledge graph completion model based on contrastive learning and relation enhancement method

Cited by 21 publications

References 34 publications

Knowledge Graph Link Prediction via Hyperbolic Attenuated Attention Networks

Knowledge Graph Link Prediction via Hyperbolic Attenuated Attention Networks

Enhanced Word-Unit Broad Learning System With Sememes

Knowledge Graph Completion Based on Entity Descriptions in Hyperbolic Space

Contact Info

Product

Resources

About