Search

Knowledge graph completion is a process of reasoning new triples based on existing entities and relations in knowledge graph. The existing methods usually use the encoder-decoder framework. Encoder uses graph convolutional neural network to get the embeddings of entities and relations. Decoder calculates the score of each tail entity according to the embeddings of the entities and relations. The tail entity with the highest score is the inference result. Decoder inferences triples independently, without consideration of graph information. Therefore, this paper proposes a graph completion algorithm based on contrastive learning. This paper adds a multi-view contrastive learning framework into the model to constrain the embedded information at graph level. The comparison of multiple views in the model constructs different distribution spaces for relations. Different distributions of relations fit each other, which is more suitable for completion tasks. Contrastive learning constraints the embedding vectors of entity and subgraph and enhahces peroformance of the task. Experiments are carried out on two datasets. The results show that MRR is improved by 12.6% over method A2N and 0.8% over InteractE on FB15k-237 dataset, and 7.3% over A2N and 4.3% over InteractE on WN18RR dataset. Experimental results demonstrate that this model outperforms other completion methods.

This research article outlines a study that examines the creation of a comprehensive knowledge graph specifically designed for the domain of urban agriculture. The research centers on the acquisition, synthesis, and arrangement of pertinent information from various origins in order to establish a specialized knowledge graph tailored for urban agricultural systems. The graph depicts the interrelationships and attributes of various entities, including urban farms, crops, farming methods, environmental factors, and economic elements. Moreover, this study investigates the efficacy of different graph embedding methodologies in the domain of urban agriculture. The aforementioned techniques are utilized in the context of the urban agriculture knowledge graph in order to extract significant representations of entities and their relationships. The primary objective of the experimental study is to investigate and reveal semantic relationships, patterns, and predictions that have the potential to improve decision-making processes and optimize practices in the field of urban agriculture. The results of this study make a significant contribution to the existing body of knowledge in the area of urban agriculture. Additionally, they offer valuable insights into the potential uses of graph embedding techniques within this field.

In real-world scenarios, obtaining the relationships between nodes is often challenging, resulting in incomplete network topology. This limitation significantly reduces the applicability of community detection methods, particularly neighborhood aggregation-based approaches, on structurally incomplete networks. Therefore, in this situation, it is crucial to obtain meaningful community information from the limited network structure. To address this challenge, the LPGSE algorithm was designed and implemented, which includes four parts: link prediction, structure observation, network estimation, and community partitioning. LPGSE demonstrated its performance in community detection in structurally incomplete networks with 10% missing edges on multiple datasets. Compared with traditional community detection algorithms, LPGSE achieved improvements in NMI and ARI metrics of 1.5781% to 29.0780% and 0.4332% to 31.9820%, respectively. Compared with similar community detection algorithms for structurally incomplete networks, LPGSE also outperformed other algorithms on all datasets. In addition, different edge-missing ratio settings were also attempted, and the performance of different algorithms in these situations was compared and analyzed. The results showed that the algorithm can still maintain high accuracy and stability in community detection across different edge-missing ratios. [ABSTRACT FROM AUTHOR]

The current industrial knowledge graph often faces the challenge of data sparsity, which can significantly impact its effectiveness and reliability in daily operational processes. To address this challenge and ensure the integrity of the knowledge graph, we propose a novel method for link prediction that leverages both entity descriptions and neighborhood structure information. Specifically, our method uses BERT pre-training to obtain meaningful embeddings from entity descriptions and the R-GCN model to capture the structural patterns within neighborhoods. Additionally, a CNN is employed to fuse and decode these two types of representations, ensuring high accuracy in predicting missing links. We have evaluated our method on publicly available datasets, and the experimental results show its superiority over baseline models. Furthermore, when tested on the SEFD dataset for steel fault diagnosis, our method effectively completes the knowledge graph for this industrial domain. [ABSTRACT FROM AUTHOR]

The identification of drug-target interactions (DTI) is a valuable step in the drug discovery and repositioning process. However, traditional laboratory experiments are time-consuming and expensive. Computational methods have streamlined research to determine DTIs. The application of deep learning methods has significantly improved the prediction performance for DTIs. Modern deep learning methods can leverage multiple sources of information, including sequence data that contains biological structural information, and interaction data. While useful, these methods cannot be effectively applied to each type of information individually (e.g., chemical structure and interaction network) and do not take into account the specificity of DTI data such as low- or zero-interaction biological entities. To overcome these limitations, we propose a method called MFA-DTI (Multi-feature Fusion Adopted framework for DTI). MFA-DTI consists of three modules: an interaction graph learning module that processes the interaction network to generate interaction vectors, a chemical structure learning module that extracts features from the chemical structure, and a fusion module that combines these features for the final prediction. To validate the performance of MFA-DTI, we conducted experiments on six public datasets under different settings. The results indicate that the proposed method is highly effective in various settings and outperforms state-of-the-art methods. • A novel DTI prediction model that explores both the graph and structural features of drugs and proteins is proposed. • A graph processing algorithm for biological networks is designed. • A module for generating sequence structure vectors is developed. • The results on 6 public datasets show it outperforms existing methods in all observed metrics. [ABSTRACT FROM AUTHOR]

The idea of predicting links in multiplex networks has gained increasing interest in recent years. In this paper, we propose a comprehensive framework which benefits from the structural information of auxiliary layers to predict links on a target layer of multiplex networks. Specifically, we assume that the likelihood of the existence of a link between two nodes is determined by the contributions from both the nodes' neighbors on the target layer and their counterparts' neighbors on a manually network generated by auxiliary layers. The final likelihood matrix is acquired by an iterative algorithm. In addition, we show advantages of our methods for predicting links on sparse and dense networks as well as on networks with assortative and disassortative structural layers. The effectiveness of the proposed methods are evaluated through extensive experiments on real-world multiplex networks. [ABSTRACT FROM AUTHOR]

In the realm of graph representation learning, Graph Neural Networks (GNNs) have demonstrated exceptional efficacy across diverse tasks. Typically, GNNs employ message-passing schemes to disseminate node features along graph structures, culminating in learned graph representations. However, their heavy reliance on smoothed node features over graph structures, coupled with limited expressiveness in the presence of node attributes, often constrains link prediction performance. To surmount this challenge, we propose GTLP, a Graph Transformer based link prediction framework. GTLP integrates unsupervised GNNs and structure encoding, enabling a holistic consideration of both topological structures and node features. This approach preserves critical node location and role information, enhancing the model’s expressiveness. By introducing the Graph Transformer model, GTLP adeptly incorporates neighbor information, refining embedding quality and bolstering the model’s learning and generalization capabilities. Notably, our method exhibits superior scalability, accommodating diverse techniques for information extraction, embedding learning, and sampling. Experimental results underscore GTLP’s state-of-the-art performance, outpacing various baselines across five real-world datasets. [ABSTRACT FROM AUTHOR]

Link prediction is one of the most essential and crucial tasks in complex network research since it seeks to forecast missing links in a network based on current ones. This problem has applications in a variety of scientific disciplines, including social network research, recommendation systems, and biological networks. In previous work, link prediction has been solved through different methods such as path, social theory, topology, and similarity-based. The main issue is that path-based methods ignore topological features, while structure-based methods also fail to combine the path and structured-based features. As a result, a new technique based on the shortest path and topological features' has been developed. The method uses both local and global similarity indices to measure the similarity. Extensive experiments on real-world datasets from a variety of domains are utilized to empirically test and compare the proposed framework to many state-of-the-art prediction techniques. Over 100 iterations, the collected data showed that the proposed method improved on the other methods in terms of accuracy. SI and AA, among the existing state-of-the-art algorithms, fared best with an AUC value of 82%, while the proposed method has an AUC value of 84%. [ABSTRACT FROM AUTHOR]

To study the evolution process and community structure of e-commerce networks, this paper uses AN improved stochastic block model (sbm) link prediction algorithm. since the degree distribution among blocks in the original sbm model is binomial, to make the degree distribution among blocks follow the power law distribution in the stochastic block model, this paper introduces the degree attenuation parameter. aiming AT the assumption that the connection between nodes depends only on the block to which nodes belong in the original sbm model, to make the degree distribution closer to the real network, the paper introduces the degree control parameter. based on this, the paper proposes AN optimized random block model, and use the alibaba taobao data set to verify the proposed algorithm. the results show that the accuracy of the proposed algorithm is higher than the sbm, the degree-corrected stochastic block model (dcsbm) and the hierarchical structure model (hbm). it shows that the improved algorithm can describe the community structure of the e-commerce network well and find the missing link in the network accurately. [ABSTRACT FROM AUTHOR]

In the process of knowledge inference, with the increase of the length of the inference path, the action space of the node increases sharply, which makes the inference difficulty continue to increase. A Knowledge Reasoning Method Of Hierarchical Reinforcement Learning (MutiAg-HRL) is proposed to reduce the size of action space in the reasoning process. MutiAg-HRL invokes high-level agents to perform rough reasoning on the relationships in the knowledge graph, and determines the approximate location of the target entity by calculating the similarity between the next step relationship and the given query relationship. According to the relationship given by the high-level agent, the low-level agents are guided to conduct detailed reasoning and select the next action. The model also constructs an interactive reward mechanism to reward the relationship between the two agents and the choice of actions in time to prevent the problem of sparse reward in the model. To verify the effectiveness of the proposed method, experiments were carried out on FB15K-237 and NELL-995 datasets. The experimental results were compared with those of 11mainstream methods such as TransE, MINERVA and HRL. The average value of the MutiAg-HRL method on the link prediction task Hits@k was increased by 1.85%. MRR increases by an average of 2%. [ABSTRACT FROM AUTHOR]

This paper presents a link prediction algorithm for dynamic social networks based on B+ trees. The authors pointed out the need for precise link prediction in social network analysis and asserted that current methods often produce inaccurate results as they are unable to account the social networks' dynamic nature. To circumvent this, a B+ tree-inspired community-based link prediction (BT-LPD) algorithm is proposed, which efficiently stores and retrieves the node information and associations and permits rapid queries of potential links. First, a community discovery technique for dynamic social networks inspired by B+ trees is described. The proposed method predicts missing links using community data and an approximation of influence flow. The performance is then evaluated using actual datasets and a number of cutting-edge methodologies. Findings show that the suggested BT-LPD technique outperforms the compared alternatives in terms of accuracy in dynamic social networks. It has important implications for social network analysis, recommender systems, and other applications that rely on accurate link prediction. The study introduces a novel and practical technique for link prediction in dynamic social networks. [ABSTRACT FROM AUTHOR]

Provenance research in museums has traditionally been reactive and focused on singular objects with dubious histories, such as colonial-era acquisitions, Nazi-looted art, and objects with active ownership claims; the 'crimes' we expect to see. But what if what we think we know prevents us from seeing the bigger picture within and across museum collections? We argue that a machine-learning approach to provenance could allow the detection of broader patterns of unethical or even criminal behaviour that are embedded in the relationships underpinning museum collections. To demonstrate the potential of a machine-learning approach, we present a computer-assisted model that predicts plausible patterns and connections, 'leads' or 'hot tips', derived from a dataset of unstructured texts concerning the antiquities trade. Preliminary results have revealed what may have been a multi-decade scheme involving the donation of low-value Latin American antiquities to museums as a form of 'reputation laundering' potentially in advance of criminal fraud. We believe that such patterns could not be identified by an approach to museum provenance that is restricted to known problems within individual institution, demonstrating the need for innovative provenance tools and approaches that consider the complex networks within which museum objects exist. [ABSTRACT FROM AUTHOR]

Knowledge graph generally has the problem of incomplete knowledge, which makes link prediction an important research content of knowledge graph. Existing models only focus on the embedding representation of triples. On the one hand, in terms of model input, only the embedding representation of entities and relations is randomly initialized, and the description information of entities and relations is not incorporated, which will lack semantic information; on the other hand, in decoding, the influence of the structural features of the triplet itself on the link prediction results is ignored. Aiming at the above problems, this paper proposes a knowledge graph link prediction model BFGAT (graph attention network link prediction based on fusion of description information and structural features) that integrates description information and structural features. The BFGAT model uses the BERT pretraining model to encode the description information of entities and relations, and integrates the description information into the embedding representation of entities and relations to solve the problem of missing semantic information. In the coding process, graph attention mechanism is used to aggregate the information of adjacent nodes to solve the problem that the target node can obtain more information. The embedding representation of triples is spliced into a matrix in the decoding process, using a method based on CNN convolution pooling to solve the problem of triple structural features. The model is subjected to detailed experiments on the public datasets FB15k-237 and WN18RR, and the experiments show that the BFGAT model can effectively improve the effect of knowledge graph link prediction.

Many real-world problems can be modelled in the form of complex networks. Social networks such as research collaboration networks and facebook, biological neural networks such as human brains, biomedical networks such as drug-target interactions and protein-protein interactions, technological networks such as telephone networks, transportation networks and power grids are a few examples of complex networks. Any complex system with entities and interactions existing between the entities can be modelled as a graph mathematically, with nodes representing entities and edges reflecting interactions. In numerous real-world circumstances, interactions are not confined to pair of entities. Majority of these intricate systems inherently possess hypergraph structures, characterized by interactions that extend beyond pairwise connections. Existing studies often transform complex interactions at a higher level into pairwise interactions and subsequently analyze them. This conversion frequently leads to both the loss of information and the inability to reconstruct the original hypergraph from the transformed network with pairwise interactions. One of the most essential tasks that can be performed on these graphs is Link Prediction (LP), which is the task of predicting future edges (links) in a graph. LP in graphs is well investigated. This article presents a novel methodology for predicting links in hypergraphs. Unlike conventional approaches that transform hypergraphs into graphs with pairwise interactions, the proposed method directly leverages the inherent structure of hypergraphs in predicting future interaction between a pair of nodes. This is motivated by the fact that hypergraphs enable the depiction of intricate higher-order relationships through hyperlinks, enhancing their representation. Their capacity to capture complex structural patterns improves predictive capabilities. Node neighborhoods within hypergraphs offer a comprehensive framework for LP, where hyperlinks simplify interactions between nodes across cliques. We propose a novel method of Link Prediction in Hypergraphs (LPH) to predict interactions within hypergraphs, maintaining their original structure without conversion to graphs, thus preserving information integrity. The proposed approach LPH extends local similarity measures like Common Neighbors, Jaccard Coefficient, Adamic Adar, and Resource Allocation, along with a global measure, Katz index, to hypergraphs. LPH’s effectiveness is assessed on six benchmark hyper-networks, employing evaluation metrics such as Area under ROC curve, Precision, and F1-score. The proposed measures of LP on hypergraphs resulted in an average enhancement of 10% in terms of Area under ROC curve compared to contemporary as well as conventional measures. Additionally, there is an average improvement of 70% in precision and around 50% in F1-score. This methodology presents a promising avenue for predicting pairwise interactions within hypergraphs while retaining their inherent structural complexity as well as information integrity.

In the link prediction problem a relevant algorithm running over a network attempts to determine whether a link between two nodes will exist in the future, given that it is not present at the moment. Most link prediction algorithms take into account the structure of the network on which they are applied and based on this, they attempt to predict the existence or not of future new edges in the network. However, many of them are quite standardized, applying the same concept and parametrization to all networks, thus not always achieving good results in every different network structure. Algorithms based on Graph Neural Networks (GNNs) are more adaptive to any network structure but they do not give appreciable results when the only information available is the network structure. In this paper, we propose a new approach to this problem that approximates the structure of a complex network by allowing adjusted weight to this network structure to create additional information, which we can embed into effective algorithms such as local and superposed random walk link prediction. To achieve this goal, we use well-known kernel functions such as Sigmoids, in which we fit their parameters appropriately by a genetic algorithm to achieve the best possible approximation. To demonstrate the effectiveness of our proposed method we have compared our prediction method results based on precision, AUC and AUPR on eleven selected networks of different structures and properties with seven well-known link prediction algorithms and one more utilizing GNNs. In every case, we have improved the results of random walk algorithms and in most cases we achieved better results from all employed benchmark algorithms.

Providing suitable flight recommendations to passengers is one of the essential requirements for ensuring customer satisfaction and maintaining a strong relationship with them in airline companies. Determining the appropriateness of a flight for a passenger is a complex issue that can depend on various factors. Factors such as individual preferences, flight quality, and the possibility of flight cancellations or delays need to be simultaneously considered in the process of making appropriate flight recommendations. Additionally, the vast amount of flight data adds to the complexity of this issue. In this article, a personalized flight recommender system is presented to address these challenges. The proposed method utilizes a link prediction strategy to model user profiles and habits, limiting the set of feasible recommendations. Furthermore, a Convolutional Neural Network (CNN) is employed to predict the likelihood of flight cancellations or delays, enabling the system to provide passengers with recommendations that maximize their satisfaction based on this information combined with flight features. To reduce the complexity of handling large flight datasets and increase processing speed, the proposed approach utilizes clustering. In this technique, data is distributed into a set of clusters using the K-Means algorithm, and the recommendation process is based on the cluster with the least distance to the user’s features. The performance of the proposed method was tested using real flight data. The experiments evaluated the model’s accuracy in predicting flight delays/cancellations and the accuracy of the recommendations it provides. The results demonstrated that the CNN model employed in the proposed method achieved an average accuracy of 95.13% in predicting flight delays/cancellations, showing at least a 2.4% improvement over the compared methods. Additionally, the proposed recommender system reported an accuracy value of 72.31%, surpassing the compared works by 15.6% and indicating its favorable performance in providing accurate recommendations.

Knowledge Graphs have demonstrated a real advantage in knowledge representation, leveraging graphs NoSQL structures and schema-less technology, which offers superior comprehension, knowledge representation, interpretation, and reasoning. The problem is that current methods for Knowledge Graph embedding rely on the topology of the graph, and essential information about entities and relations has not been fully employed, failing to utilize the graph’s ontology to limit the spurious growth of edges, leading to inaccurate, misleading, and fabricated knowledge. This research aims to establish a method to restrict the spurious growth of host graph by imposing an upper bound on edge embedding using the claim’s and the host’s ontology graph. Through this research, a claim-ontology signature artifact was designed to facilitate open-environment KG completion. This artifact establishes the upper bound for the type of edges predicted by the link prediction algorithm, thus preventing the spurious growth of edges within the Knowledge Graph. Furthermore, the artifact was evaluated in the context of three use cases: host-guided embedding, claim-guided embedding, and topic-guided embedding, using a quantitative framework for design science evaluation. The main finding is that the spurious growth of edges can be limited by imposing an upper bound on the possible edge embedding using the claim’s graph and the host ontology graph. A secondary finding is that the artifact could serve as an instrument to manage the ontology-topology tradeoff in Knowledge Graphs.

Link prediction in complex networks is a critical process aimed at uncovering hidden or potential connections among nodes. This technique is widely utilized in areas such as knowledge graphs. Current Graph Neural Networks (GNNs) often focus exclusively on determining whether nodes are connected or assessing the strength of these links by leveraging node attributes. They typically use network structure and attributes to develop node representations through neighborhood aggregation. However, these methods often overlook the intrinsic importance of the links themselves. This paper thoroughly examines the significance of link value based on network structure and introduces an innovative approach for estimating this value, and proposes a method that incorporates link value into both the formulation and training of a link prediction graph attention network. This integration not only boosts the accuracy of link predictions but also provides a theoretical basis for understanding the prediction results. We conducted extensive experiments in link prediction employing widely recognized benchmark datasets. The findings reveal that our proposed framework for link prediction exhibits commendable performance and generalization capabilities, and overall performance improved by an average of 1.2%, thereby establishing it as an effective baseline model.

Complex networks are widely used to represent an abundance of real-world relations ranging from social networks to brain networks. Inferring missing links or predicting future ones based on the currently observed network is known as the link prediction task. Recent network embedding based link prediction algorithms have demonstrated ground-breaking performance on link prediction accuracy. Those algorithms usually apply node attributes as the initial feature input to accelerate the convergence speed during the training process. However, they do not take full advantage of node feature information. In this paper, besides applying feature attributes as the initial input, we make better utilization of node attribute information by building attributable networks and plugging attributable networks into some typical link prediction algorithms and name this algorithm Attributive Graph Enhanced Embedding (AGEE). AGEE is able to automatically learn the weighting trades-off between the structure and the attributive networks. Numerical experiments show that AGEE can improve the link prediction accuracy by around 3% compared with SEAL, Variational Graph AutoEncoder (VGAE), and node2vec.

LncRNAs are an essential type of non-coding RNAs, which have been reported to be involved in various human pathological conditions. Increasing evidence suggests that drugs can regulate lncRNAs expression, which makes it possible to develop lncRNAs as therapeutic targets. Thus, developing in-silico methods to predict lncRNA-drug associations (LDAs) is a critical step for developing lncRNA-based therapies. In this study, we predict LDAs by using graph convolutional networks (GCN) and graph attention networks (GAT) based on lncRNA and drug similarity networks. Results show that our proposed method achieves good performance (average AUCs > 0.92) on five datasets. In addition, case studies and KEGG functional enrichment analysis further prove that the model can effectively identify novel LDAs. On the whole, this study provides a deep learning-based framework for predicting novel LDAs, which will accelerate the lncRNA-targeted drug development process.

Graph representation learning is a fundamental technique for machine learning (ML) on complex networks. Given an input network, these methods represent the vertices by low-dimensional real-valued vectors. These vectors can be used for a multitude of downstream ML tasks. We study one of the most important such task, link prediction. Much of the recent literature on graph representation learning has shown remarkable success in link prediction. On closer investigation, we observe that the performance is measured by the AUC (area under the curve), which suffers biases. Since the ground truth in link prediction is sparse, we design a vertex-centric measure of performance, called the VCMPR@k plots. Under this measure, we show that link predictors using graph representations show poor scores. Despite having extremely high AUC scores, the predictors miss much of the ground truth. We identify a mathematical connection between this performance, the sparsity of the ground truth, and the low-dimensional geometry of the node embeddings. Under a formal theoretical framework, we prove that low-dimensional vectors cannot capture sparse ground truth using dot product similarities (the standard practice in the literature). Our results call into question existing results on link prediction and pose a significant scientific challenge for graph representation learning. The VCMPR plots identify specific scientific challenges for link prediction using low-dimensional node embeddings. [ABSTRACT FROM AUTHOR]

This study presents a novel approach that synergizes community detection algorithms with various Graph Neural Network (GNN) models to bolster link prediction in scientific literature networks. By integrating the Louvain community detection algorithm into our GNN frameworks, we consistently enhanced the performance across all models tested. For example, integrating the Louvain model with the GAT model resulted in an AUC score increase from 0.777 to 0.823, exemplifying the typical improvements observed. Similar gains were noted when the Louvain model was paired with other GNN architectures, confirming the robustness and effectiveness of incorporating community-level insights. This consistent increase in performance—reflected in our extensive experimentation on bipartite graphs of scientific collaborations and citations—highlights the synergistic potential of combining community detection with GNNs to overcome common link prediction challenges such as scalability and resolution limits. Our findings advocate for the integration of community structures as a significant step forward in the predictive accuracy of network science models, offering a comprehensive understanding of scientific collaboration patterns through the lens of advanced machine learning techniques. [ABSTRACT FROM AUTHOR]

The interaction between human microbes and drugs can significantly impact human physiological functions. It is crucial to identify potential microbe-drug associations (MDAs) before drug administration. However, conventional biological experiments to predict MDAs are plagued by drawbacks such as time-consuming, high costs, and potential risks. On the contrary, computational approaches can speed up the screening of MDAs at a low cost. Most computational models usually use a drug similarity matrix as the initial feature representation of drugs and stack the graph neural network layers to extract the features of network nodes. However, different calculation methods result in distinct similarity matrices, and message passing in graph neural networks (GNNs) induces phenomena of over-smoothing and over-squashing, thereby impacting the performance of the model. To address these issues, we proposed a novel graph representation learning model, dual-modal graph learning for microbe-drug association prediction (DMGL-MDA). It comprises a dual-modal embedding module, a bipartite graph network embedding module, and a predictor module. To assess the performance of DMGL-MDA, we compared it against state-of-the-art methods using two benchmark datasets. Through cross-validation, we illustrated the superiority of DMGL-MDA. Furthermore, we conducted ablation experiments and case studies to validate the effective performance of the model. • DMGL-MDA employs a dual-modal strategy, integrating fingerprints and graph features to obtain a comprehensive representation. • DMGL-MDA integrates local & global information to alleviate over-smoothing or over-squashing phenomena observed in GNNs. • DMGL-MDA outperforms state-of-the-art approaches and it achieves the best performance on average. [ABSTRACT FROM AUTHOR]

The knowledge graph in the recommendation system plays a vital role in the recommendation effect of the system, and the knowledge representation in the graph becomes a key factor affecting the recommendation system, which has become one of the current research hotspots. This paper proposed a node2vec-based knowledge representation node2 vec-side based on the traditional node2vec model by adding relational representation and diversifing wandering strategy to the structural characteristics of the knowledge graph in recommendation system, which combined with the knowledge graph network structure of recommendation system to explore the potential association relationship between nodes of large-scale recommendation entities, reduced the complexity of the representation and improved interpretability. After time complexity analysis, it could be seen that the proposed knowledge representation is lower than Trans series and RGCN in terms of complexity. Link prediction experiments were conducted on the traditional knowledge graph datasets FB15K, WN18, and recommendation domain datasets MovieLens-1M, Book-Crossing, Last. FM respectively. The experimental results show that on the MovieLens-1M dataset, hits @ 10 improves 5.5%12.1% and MRR improves 0. 090. 24, respectively. On the Book-Crossing dataset, hits @ 10 improves 3.5%-20.6%, and MRR improves 0.04-0.24 on average, respectively. And on the Last. FM dataset, hits@1 improves 0.3%-8.5% and MRR improves 0.04-0. 16 on average. It is better than the existing algorithms and verifies the effectiveness of the proposed method. [ABSTRACT FROM AUTHOR]

Link prediction aims to anticipate the probability of a future connection between two nodes in a given network based on their previous interactions and the network structure. Link prediction is a rapidly evolving field of research that has attracted interest from physicists and computer scientists. Over the years, numerous methods have been developed for link prediction, encompassing similarity-based indices, machine learning techniques, and more. While existing surveys have covered link prediction research until 2020, there has been a substantial surge in research activities in recent years, particularly between 2021 and 2023. This increased interest underscores the pressing need to comprehensively explore the latest advancements and approaches in link prediction. We analyse and present the most notable research from 2018 to 2023. Our goal is to offer a comprehensive overview of the recent developments in the field. Besides summarizing and presenting previous experimental results, our survey offers a comprehensive analysis highlighting the strengths and limitations of various link prediction methods. [ABSTRACT FROM AUTHOR]