Search

With the rapid advancement of scientific research, collaboration in this area is becoming increasingly important. One of the major challenges is the link prediction problem for research collaboration. Recently, learning-based link prediction methods have received much attention. However, most of these studies have solely concentrated on exploiting a single network and its topology features for prediction, and ignore other factors that may influence link formation. To address this issue, in this paper we propose a link prediction model based on multi-network representation learning. Specifically, we develop new features based on the author's institutions and published papers, and three networks incorporating these features are modeled. Then, the network representation method based on joint training is proposed to embed the networks in a low-dimensional space. Finally, the authors' feature vectors are combined in finer granularity, and collaboration prediction is performed in a supervised manner. The performance of our model is evaluated by comparing it with other link prediction methods on a real-world dataset, and the experimental results show the effectiveness of our model. [ABSTRACT FROM AUTHOR]

Link prediction is an important task in social network analysis, with the goal of evaluating the possibility of forming links between node pairs. In this paper, we introduce compressed sensing technology into the link prediction task and propose a new link prediction method. Here, we transform the link prediction problem into a signal processing process. The idea is to take the links observed in social networks as the encoding result of compressed sensing signals, and then decode the encoded signal using the reconstruction algorithm to obtain more complete network information. We propose a sparse representation algorithm based on K-Singular Value Decomposition (K-SVD), which can integrate network topologies and node attribute information. On this basis, we construct the measurement matrix conforming to the Restricted Isometry Property (RIP), and further develop the Orthogonal Matching Pursuit (OMP) algorithm to handle the reconstruction issues. Experimental results on ten different types of real social networks show that the proposed method outperforms other state-of-the-art link prediction algorithms in terms of both accuracy and universality. In particular, it still performs stably when the network is sparse. [ABSTRACT FROM AUTHOR]

Link prediction is a crucial area of study within complex networks research. Mapping nodes to low-dimensional vectors through network embeddings is a vital technique for link prediction. Most of the existing methods employ "node–edge"-structured networks to model the data and learn node embeddings. In this paper, we initially introduce the Clique structure to enhance the existing model and investigate the impact of introducing two Clique structures (LECON: Learning Embedding based on Clique Of the Network) and nine motifs (LEMON: Learning Embedding based on Motif Of the Network), respectively, on experimental performance. Subsequently, we introduce a hypergraph to model the network and reconfigure the network by mapping hypermotifs to two structures: open hypermotif and closed hypermotif, respectively. Then, we introduce hypermotifs as supernodes to capture the structural similarity between nodes in the network (HMRLH: HyperMotif Representation Learning on Hypergraph). After that, taking into account the connectivity and structural similarity of the involved nodes, we propose the Depth and Breadth Motif Random Walk method to acquire node sequences. We then apply this method to the LEMON (LEMON-DB: LEMON-Depth and Breadth Motif Random Walk) and HMRLH (HMRLH-DB: HMRLH-Depth and Breadth Motif Random Walk) algorithms. The experimental results on four different datasets indicate that, compared with the LEMON method, the LECON method improves experimental performance while reducing time complexity. The HMRLH method, utilizing hypernetwork modeling, proves more effective in extracting node features. The LEMON-DB and HMRLH-DB methods, incorporating new random walk approaches, outperform the original methods in the field of link prediction. Compared with state-of-the-art baselines, the proposed approach in this paper effectively enhances link prediction accuracy, demonstrating a certain level of superiority. [ABSTRACT FROM AUTHOR]

Knowledge hypergraphs (KHG) are knowledge graph of hypergraph structure. KHG link prediction aims to predict the missing relations through the known entities and relations. However, HypE, the existing optimal KHG link prediction method based on embedding model considers the location information when embedding entities, but ignores the differences in the contributions of different entities when embedding relations. And the information of the entity convolution vector is insufficient. Relation embeddings consider the entity contribution and supply the information of entity embedding, which can greatly improve the prediction ability of model. Therefore, link prediction based on attention and convolution network (LPACN) is proposed. The improved attention mechanism is applied to merging entity attention information into relation embeddings. And the number information of neighboring entities in the same tuple is integrated into the convolution network, which further supplies the information of entity convolution embedding. For the gradient vanishing problem of LPACN, the improved ResidualNet is integrated into LPACN, and the multilayer perceptron (MLP) is used to improve the nonlinear learning ability of model. The improved algorithm LPACN+ is obtained. Extensive experiments on real datasets validate that LPACN is better than the Baselines.

Purposes The brain functional network construction method of resting state functional magnetic resonance imaging (rs-fMRI) has been relatively mature, but the fMRI signal acquisition process is affected by acquisition equipment, subjects’ own reasons, noise, and other factors. The traditional rs-fMRI brain network construction method may have false links or missing edges, the network expression accuracy and the stability of repeated measurement needs to be further improved. Methods In order to solve above problems, on the basis of three local information indexes and four node definition strategies, the link prediction algorithm is applied to brain function network reconstruction, and two groups of test data are used to analyze the reliability of the reconstructed network topology index. Findings The experimental results show that the reconstructed network improves the stability of repeated measurement. In addition, the inter group difference of brain network in patients with Alzheimer's disease (AD) is analyzed by using the reconstructed network. The results show that there are significant inter group differences in the reconstructed network, which is consistent with the existing research results. This study provides a certain reference for the technology of constructing a stable brain function network.

The study’s findings showed that link prediction utilizing the similarity learning model in dynamic networks (LSDN) performed better than other learning techniques including neural network learning and decision tree learning in terms of the three criteria of accuracy, coverage, and efficiency., Compared to the random forest approach, the LSDN learning algorithm’s link prediction accuracy increased from 97% to 99%. The proposed method’s use of oversampling, which improved link prediction accuracy, was the cause of the improvement in area under the curve (AUC). To bring the ratio of the classes closer together, the suggested strategy attempted to produce more samples from the minority class. In addition, similarity criteria were chosen utilizing feature selection techniques based on correlation that had a strong link with classes. This technique decreased over-fitting and improved the suggested method’s test data generalizability. Based on the three criteria (accuracy, coverage, and efficiency), the research’s findings demonstrated that link prediction utilizing the similarity LSDN outperformed other learning techniques including neural network learning and decision tree learning. Compared to the random forest algorithm, the LSDN algorithm’s link prediction accuracy increased from 97% to 99%. The oversampling in the suggested strategy, which increased link prediction accuracy, is what caused the increase in AUC. To bring the ratio of the classes closer together, the suggested strategy attempted to produce more samples from the minority class. In addition, similarity criteria were chosen utilizing feature selection techniques based on correlation that had a strong link with classes. This technique decreased over-fitting and improved the suggested method’s test data generalizability.

Abstract Influence maximization (IM) is the task of finding the most important nodes in order to maximize the spread of influence or information on a network. This task is typically studied on static or temporal networks where the complete topology of the graph is known. In practice, however, the seed nodes must be selected before observing the future evolution of the network. In this work, we consider this realistic ex ante setting where p time steps of the network have been observed before selecting the seed nodes. Then the influence is calculated after the network continues to evolve for a total of $$T>p$$ T > p time steps. We address this problem by using statistical, non-negative matrix factorization and graph neural networks link prediction algorithms to predict the future evolution of the network, and then apply existing influence maximization algorithms on the predicted networks. Additionally, the output of the link prediction methods can be used to construct novel IM algorithms. We apply the proposed methods to eight real-world and synthetic networks to compare their performance using the susceptible-infected (SI) diffusion model. We demonstrate that it is possible to construct quality seed sets in the ex ante setting as we achieve influence spread within 87% of the optimal spread on seven of eight network. In many settings, choosing seed nodes based only historical edges provides results comparable to the results treating the future graph snapshots as known. The proposed heuristics based on the link prediction model are also some of the best-performing methods. These findings indicate that, for these eight networks under the SI model, the latent process which determines the most influential nodes may not have large temporal variation. Thus, knowing the future status of the network is not necessary to obtain good results for ex ante IM.

One of these challenges in the analysis of social networks is the problem of link prediction. The purpose of this problem is to find links that have not yet been observed, but may exist in the future. There are many solutions for link prediction on monoplex networks. However, many real social networks model communication in multiple layers, which are known as multilayer social networks. A solution for multilayer networks involves taking into account the information of all layers to make predictions for a target layer. Among the existing solutions, local random walk has been confirmed as an efficient technique for link prediction in monoplex networks, but this technique is inefficient for link prediction in multilayer networks due to computational complexity. In order to address this issue, in this paper we propose Extended Neighborhood based Local Random Walk (ENLRW) for link prediction in multilayer networks. ENLRW is an extended version of the classical local random walk technique in which the nearest neighbors are considered based on the extended neighborhood concept. ENLRW calculates the similarity between vertices by integrating several different metrics through reliable paths that include intra-layer and inter-layer information. Besides, ENLRW considers vertex influence as a similarity metric to provide an effective reliable biased random walk. The results of the simulations show that the use of different inter-layer and intra-layer information as well as the local random walk configuration with extended neighborhood provides a trade-off between precision and complexity. Specifically, ENLRW improves the average precision by 3.1% compared to the best available state-of-the-art method.

The information age needs to change the traditional physical education teaching mode, set up modern teaching concepts, and pay attention to the cultivation of strong physical talents. This paper combines the ADDIE model and MOOC education platform to construct a hybrid teaching model for college sports, and gives the teaching design of the three phases of college sports hybrid teaching before, during and after class. Based on the GCN model and the similarity between MOOC sports teaching resources, a sports teaching resource recommendation model based on link prediction was proposed. Relevant research subjects and questionnaires were selected, and a blended teaching experiment was designed. Empirical analysis was conducted using the method of comparison of teaching experiments. The results show that the NDCG metrics of sports teaching resources based on the GCN model improved by 0.99%, 1.30%, and 1.16% on average compared with the CF-User algorithm on the Gowalla, Yelp2018, and Amazon-Book datasets. After going through the sports blended teaching experiment, the students in the experimental group improved their average scores on exercise attitude by 1.5 to 5 points, and their interest in learning increased by about 1 point. The blended teaching mode of combining MOOC platform and college sports teaching in the information age has changed students’ learning attitude to a certain extent, which can effectively enhance the quality of college sports teaching and improve the cultivation of college talents’ physical quality.

While Knowledge Graphs (KGs) have long been used as valuable sources of structured knowledge, in recent years, KG embeddings have become a popular way of deriving numeric vector representations from them, for instance, to support knowledge graph completion and similarity search. This study surveys advances as well as open challenges and opportunities in this area. For instance, the most prominent embedding models focus primarily on structural information. However, there has been notable progress in incorporating further aspects, such as semantics, multi-modal, temporal, and multilingual features. Most embedding techniques are assessed using human-curated benchmark datasets for the task of link prediction, neglecting other important real-world KG applications. Many approaches assume a static knowledge graph and are unable to account for dynamic changes. Additionally, KG embeddings may encode data biases and lack interpretability. Overall, this study provides an overview of promising research avenues to learn improved KG embeddings that can address a more diverse range of use cases.

Compared to black-box neural networks, logic rules express explicit knowledge, can provide human-understandable explanations for reasoning processes, and have found their wide application in knowledge graphs and other downstream tasks. As extracting rules manually from large knowledge graphs is labour-intensive and often infeasible, automated rule learning has recently attracted significant interest, and a number of approaches to rule learning for knowledge graphs have been proposed. This survey aims to provide a review of approaches and a classification of state-of-the-art systems for learning first-order logic rules over knowledge graphs. A comparative analysis of various approaches to rule learning is conducted based on rule language biases, underlying methods, and evaluation metrics. The approaches we consider include inductive logic programming (ILP)-based, statistical path generalisation, and neuro-symbolic methods. Moreover, we highlight important and promising application scenarios of rule learning, such as rule-based knowledge graph completion, fact checking, and applications in other research areas.

Recent studies have shown that aberrations or mutations of microRNAs (miRNA) can cause various diseases. Therefore, identifying disease-associated microRNAs can aid in diagnosing and treating related disorders. However, getting precise correlations through biological research is expensive and time-consuming. This research suggests a machine learning method (HNDLM) based on network internal topology data to anticipate disease-miRNA connections. HNDLM applies the recently described network embedding technique to biological networks rather than creating similar ones. According to the experimental findings, HNDLM performs better in terms of accuracy and Area Under the ROC Curve (AUC) value when compared to the traditional algorithms MIDPE, MIDP, WBSMDA, RLSMDA, Collective Prediction based on Transductive Learning (CPTL), and Human Disease-related MiRNA Prediction (HDMP). The top 30 potential miRNAs suggested by HNDLM can also be verified by earlier research in the case study. HNDLM can identify probable disease-miRNA correlations, which will aid in the investigation of illness path physiology and advance the field of bioinformatics.

The number of published academic papers has been increasing rapidly from year to year. However, this increase in publications must be linear with an emphasis on quality. To ensure that academic papers meet the required quality standard, the peer review process is necessary. The main objective of the assignment of reviewers is to find the appropriate reviewer who can conduct a review based on their field of research. However, there are potential obstacles when there is a conflict of interest in the process. This study aims to develop a method for assigning reviewers that overcomes such obstacles. Our approach involves combining the Latent Dirichlet Allocation (LDA), Classification, and Link Prediction methods. LDA is used to find topics from the research data of prospective reviewers to ensure that the assigned reviewers are well suited to the submitted article. These data were used as training data for classification using Random Forest. Finally, link prediction implemented to make reviewer recommendations. We evaluated and compared our proposed method with previous research that used cosine similarity as the last step in recommendation, using Mean Average Precision (MAP). Our proposed method achieved a MAP value of 0.87, which was an improvement compared to the previous approach. These results suggest that our approach has the potential to improve the effectiveness of academic peer review.

With the rise of multi-modal methods, multi-modal knowledge graphs have become a better choice for storing human knowledge. However, knowledge graphs often suffer from the problem of incompleteness due to the infinite and constantly updating nature of knowledge, and thus the task of knowledge graph completion has been proposed. Existing multi-modal knowledge graph completion methods mostly rely on either embedding-based representations or graph neural networks, and there is still room for improvement in terms of interpretability and the ability to handle multi-hop tasks. Therefore, we propose a new method for multi-modal knowledge graph completion. Our method aims to learn multi-level graph structural features to fully explore hidden relationships within the knowledge graph and to improve reasoning accuracy. Specifically, we first use a Transformer architecture to separately learn about data representations for both the image and text modalities. Then, with the help of multimodal gating units, we filter out irrelevant information and perform feature fusion to obtain a unified encoding of knowledge representations. Furthermore, we extract multi-level path features using a width-adjustable sliding window and learn about structural feature information in the knowledge graph using graph convolutional operations. Finally, we use a scoring function to evaluate the probability of the truthfulness of encoded triplets and to complete the prediction task. To demonstrate the effectiveness of the model, we conduct experiments on two publicly available datasets, FB15K-237-IMG and WN18-IMG, and achieve improvements of 1.8 and 0.7%, respectively, in the Hits@1 metric.

Explanations in recommender systems are a requirement to improve users’ trust and experience. Traditionally, explanations in recommender systems are derived from their internal data regarding ratings, item features, and user profiles. However, this information is not available in black-box recommender systems that lack sufficient data transparency. This current work proposes a local model-agnostic, explanation-by-example method for recommender systems based on knowledge graphs to leverage this knowledge requirement. It only requires information about the interactions between users and items. Through the proper transformation of these knowledge graphs into item-based and user-based structures, link prediction techniques are applied to find similarities between the nodes and to identify explanatory items for the user’s recommendation. Experimental evaluation demonstrates that these knowledge graphs are more effective than classical content-based explanation approaches but have lower information requirements, making them more suitable for black-box recommender systems.

Abstract Background The identification of chemical–target interaction is key to pharmaceutical research and development, but the unclear materials basis and complex mechanisms of traditional medicine (TM) make it difficult, especially for low-content chemicals which are hard to test in experiments. In this research, we aim to apply the node2vec algorithm in the context of drug-herb interactions for expanding potential targets and taking advantage of molecular docking and experiments for verification. Methods Regarding the widely reported risks between cardiovascular drugs and herbs, Salvia miltiorrhiza (Danshen, DS) and Ligusticum chuanxiong (Chuanxiong, CX), which are widely used in the treatment of cardiovascular disease (CVD), and approved drugs for CVD form the new dataset as an example. Three data groups DS-drug, CX-drug, and DS-CX-drug were applied to serve as the context of drug-herb interactions for link prediction. Three types of datasets were set under three groups, containing information from chemical-target connection (CTC), chemical-chemical connection (CCC) and protein–protein interaction (PPI) in increasing steps. Five algorithms, including node2vec, were applied as comparisons. Molecular docking and pharmacological experiments were used for verification. Results Node2vec represented the best performance with average AUROC and AP values of 0.91 on the datasets “CTC, CCC, PPI”. Targets of 32 herbal chemicals were identified within 43 predicted edges of herbal chemicals and drug targets. Among them, 11 potential chemical-drug target interactions showed better binding affinity by molecular docking. Further pharmacological experiments indicated caffeic acid increased the thermal stability of the protein GGT1 and ligustilide and low-content chemical neocryptotanshinone induced mRNA change of FGF2 and MTNR1A, respectively. Conclusions The analytical framework and methods established in the study provide an important reference for researchers in discovering herb–drug interactions, alerting clinical risks, and understanding complex mechanisms of TM.

A knowledge graph is a structured semantic network designed to describe physical entities and relations in the world. A comprehensive and accurate knowledge graph is essential for tasks such as knowledge inference and recommendation systems, making link prediction a popular problem for knowledge graph completion. However, existing approaches struggle to model complex relations among entities, which severely hampers their ability to complete knowledge graphs effectively. To address this challenge, we propose a novel hierarchical multi-head attention network embedding framework, called RiQ-KGC, which integrates different-grained contextual information of knowledge graph triples and models quaternion rotation relations between entities. Furthermore, we propose a relation instantiation method for alleviating the difficulty of expressing complex relations between entities. To enhance the expressiveness of relation representation, the relation is integrated by Transformer to obtain multi-hop neighbor information, so that one relation can be embedded into different embeddings according to different entities. Experimental results on four datasets demonstrate that RiQ-KGC exhibits strong competitiveness compared to state-of-the-art models in link prediction, while the ablation experiments reveal that the proposed relation instantiation method achieves great performance.

A knowledge graph is a repository that represents a vast amount of information in the form of triplets. In the training process of completing the knowledge graph, the knowledge graph only contains positive examples, which makes reliable link prediction difficult, especially in the setting of complex relations. At the same time, current techniques that rely on distance models encapsulate entities within Euclidean space, limiting their ability to depict nuanced relationships and failing to capture their semantic importance. This research offers a unique strategy based on Gibbs sampling and connection embedding to improve the model’s competency in handling link prediction within complex relationships. Gibbs sampling is initially used to obtain high-quality negative samples. Following that, the triplet entities are mapped onto a hyperplane defined by the connection. This procedure produces complicated relationship embeddings loaded with semantic information. Through metric learning, this process produces complex relationship embeddings imbued with semantic meaning. Finally, the method’s effectiveness is demonstrated on three link prediction benchmark datasets FB15k-237, WN11RR and FB15k.

Dynamic Link Prediction (DLP) addresses the prediction of future links in evolving networks. However, accurately portraying the performance of DLP algorithms poses challenges that might impede progress in the field. Importantly, common evaluation pipelines usually calculate ranking or binary classification metrics, where the scores of observed interactions (positives) are compared with those of randomly generated ones (negatives). However, a single metric is not sufficient to fully capture the differences between DLP algorithms, and is prone to overly optimistic performance evaluation. Instead, an in-depth evaluation should reflect performance variations across different nodes, edges, and time segments. In this work, we contribute tools to perform such a comprehensive evaluation. (1) We propose Birth–Death diagrams, a simple but powerful visualization technique that illustrates the effect of time-based train–test splitting on the difficulty of DLP on a given dataset. (2) We describe an exhaustive taxonomy of negative sampling methods that can be used at evaluation time. (3) We carry out an empirical study of the effect of the different negative sampling strategies. Our comparison between heuristics and state-of-the-art memory-based methods on various real-world datasets confirms a strong effect of using different negative sampling strategies on the test area under the curve (AUC). Moreover, we conduct a visual exploration of the prediction, with additional insights on which different types of errors are prominent over time.

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.

A considerable amount of research on link prediction has recently been driven by missing relationships between knowledge graph entities and the problem of the incompleteness of knowledge graphs. Some recent studies have shown that convolutional neural networks based on knowledge embeddings are highly expressive and have good performance in link prediction. However, we found that the convolutional neural network (CNN)-based models do not handle the link between relations and entities well. For this reason, this paper proposes a link prediction model (LPM) based on feature mapping and bi-directional convolution. For the modeling of the task, an encoding layer–mapping layer–decoding layer structure is used. Among these layers, the encoding layer adopts a graph attention network to encode multi-hop triad information and obtains richer encoding of entities and relationships. The mapping layer can realize the mapping transformation between entities and relations and project the entity encoding in the space of relation encoding to capture the subtle connection between entities and relations. The decoding layer adopts bidirectional convolution to merge and decode the triples in a sequential inverse order, which makes the decoding layer model more advantageous in prediction. In addition, the decoding layer also adopts the r-drop training method to effectively reduce the distribution error generated by training between models and enhance the robustness of the model. Our experiments demonstrated the effectiveness of mapping relations, bidirectional convolution, and r-drop, and the accuracy of the proposed model showed significant improvements for each evaluation metric on two datasets, WN18RR and FB15k-237.

Link prediction for opportunistic networks faces the challenges of frequent changes in topology and complex and variable spatial-temporal information. Most existing studies focus on temporal or spatial features, ignoring ample potential information. In order to better capture the spatial-temporal correlations in the evolution of networks and explore their potential information, a link prediction method based on spatial-temporal attention and temporal convolution network (STA-TCN) is proposed. It slices opportunistic networks into discrete network snapshots. A state matrix based on topology information and attribute information is constructed to represent snapshots. Time convolutional networks and spatial-temporal attention mechanisms are employed to learn spatial-temporal information. Furthermore, to better improve link prediction performance, the proposed method converts the auto-correlation error into non-correlation error. On three real opportunistic network datasets, ITC, MIT, and Infocom06, experimental results demonstrate the superior predictive performance of the proposed method compared to baseline models, as shown by improved AUC and F1-score metrics.

Link prediction aims to identify unknown or missing connections in a network. The methods based on network structure similarity, known for their simplicity and effectiveness, have garnered widespread attention. A core metric in these methods is "proximity", which measures the similarity or linking probability between two nodes. These methods generally operate under the assumption that node pairs with higher proximity are more likely to form new connections. However, the accuracy of existing node proximity-based link prediction algorithms requires improvement. To address this, this paper introduces a Link Prediction Algorithm Based on Weighted Local and Global Closeness (LGC). This algorithm integrates the clustering coefficient to enhance prediction accuracy. A significant advantage of LGC is its dual consideration of a network's local and global features, allowing for a more precise assessment of node similarity. In experiments conducted on ten real-world datasets, the proposed LGC algorithm outperformed eight traditional link prediction methods, showing notable improvements in key evaluation metrics, namely precision and AUC. [ABSTRACT FROM AUTHOR]

Single-cell RNA-sequencing (scRNA-seq) has emerged as a powerful technique for studying gene expression patterns at the single-cell level. Inferring gene regulatory networks (GRNs) from scRNA-seq data provides insight into cellular phenotypes from the genomic level. However, the high sparsity, noise and dropout events inherent in scRNA-seq data present challenges for GRN inference. In recent years, the dramatic increase in data on experimentally validated transcription factors binding to DNA has made it possible to infer GRNs by supervised methods. In this study, we address the problem of GRN inference by framing it as a graph link prediction task. In this paper, we propose a novel framework called GNNLink, which leverages known GRNs to deduce the potential regulatory interdependencies between genes. First, we preprocess the raw scRNA-seq data. Then, we introduce a graph convolutional network-based interaction graph encoder to effectively refine gene features by capturing interdependencies between nodes in the network. Finally, the inference of GRN is obtained by performing matrix completion operation on node features. The features obtained from model training can be applied to downstream tasks such as measuring similarity and inferring causality between gene pairs. To evaluate the performance of GNNLink, we compare it with six existing GRN reconstruction methods using seven scRNA-seq datasets. These datasets encompass diverse ground truth networks, including functional interaction networks, Loss of Function/Gain of Function data, non-specific ChIP-seq data and cell-type-specific ChIP-seq data. Our experimental results demonstrate that GNNLink achieves comparable or superior performance across these datasets, showcasing its robustness and accuracy. Furthermore, we observe consistent performance across datasets of varying scales. For reproducibility, we provide the data and source code of GNNLink on our GitHub repository: https://github.com/sdesignates/GNNLink. [ABSTRACT FROM AUTHOR]

This paper proposes a simple knowledge graph embedding (KGE) framework that considers the entity type information without additional resources. The KGE is used to obtain vector representations of entities and relations by learning structured information in triples. The obtained vectors are used to predict the missing links in a knowledge graph (KG). Although many KGs contain entity type information, most of the existing methods ignored the potential of the entity type information for the link prediction task. The proposed framework, which is called entity and entity type composition representation learning (EETCRL), obtains vector representations of both entities and entity types, which are combined and used for link prediction. Experimental results on three datasets show that the EETCRL outperforms the baseline methods in most cases. Furthermore, the results obtained from tests with different model sizes show that the proposed framework can achieve high performance even with a small model size. This paper also discusses the effect of considering information about entity types on the link prediction task by analyzing the experimental results. [ABSTRACT FROM AUTHOR]

Vehicle trajectories represent an essential information source in intelligent transportation systems. Prior trajectory completion models based on Automatic Number Plate Recognition (ANPR) data have typically depended on vehicle re-identification results or road network information or have employed static knowledge graphs to integrate the two information sources. However, these methods have not taken into account the implicit temporal characteristics of trajectories in ANPR data and have neglected individual vehicle preferences. To address this void, this study proposes a Temporal Knowledge Graph-based Vehicle Trajectory Completion Model (TKG-VTC). The model implementation comprises three stages: first, ANPR data are converted into a temporal trajectory knowledge graph; second, knowledge representation learning is conducted using nontemporal relations, a biased temporal regularizer and multivector embeddings to embed the knowledge on the graph; and finally, the embedded results are employed to perform link prediction for incomplete trajectories, thereby restoring vehicle trajectories in ANPR data. Through model evaluation metrics and dimensionality reduction experiments, TKG-VTC is observed to demonstrate the best performance in completing trajectories when compared to TComplEx, TNTComplEx, and TeLM. This research introduces an innovative application of employing temporal knowledge graphs for trajectory reconstruction, which eliminates dependence on vehicle re-identification and road network information in previous methodologies. This is advantageous for enhancing the performance and dependability of vehicle trajectory data in intelligent transportation systems, as well as facilitating the implementation of trajectory prediction, demand analysis, and accident warning applications. [ABSTRACT FROM AUTHOR]

Industrial interconnection network is heterogeneous cross-border cooperation network that empower “intelligent manufacturing” and facilitate the flow of innovation elements between nodes, which in turn influence the predicted outcomes of the cooperation links. This paper proposed a link prediction method based on node innovation elements to address this issue. Firstly, it elaborated the concept and characteristics of industrial interconnection networks, and analyzed the circulation methods of innovative elements in the network. Secondly, it calculated the innovation factor quantity of network nodes from two dimensions separately. Finally, it comprehensively utilized the local information of nodes and the innovative element information of nodes to calculate the similarity between nodes. The experimental results on four industrial interconnection network datasets show that the proposed method has better predictive performance compared to other methods. [ABSTRACT FROM AUTHOR]

The rise of graph representation learning as the primary solution for many different network science tasks led to a surge of interest in the fairness of this family of methods. Link prediction, in particular, has a substantial social impact. However, link prediction algorithms tend to increase the segregation in social networks by disfavouring the links between individuals in specific demographic groups. This paper proposes a novel way to enforce fairness on graph neural networks with a fine-tuning strategy. We D rop the unfair E dges and, simultaneously, we A dapt the model's parameters to those modifications, DEA in short. We introduce two covariance-based constraints designed explicitly for the link prediction task. We use these constraints to guide the optimization process responsible for learning the new 'fair' adjacency matrix. One novelty of DEA is that we can use a discrete yet learnable adjacency matrix in our fine-tuning. We demonstrate the effectiveness of our approach on five real-world datasets and show that we can improve both the accuracy and the fairness of the link prediction tasks. In addition, we present an in-depth ablation study demonstrating that our training algorithm for the adjacency matrix can be used to improve link prediction performances during training. Finally, we compute the relevance of each component of our framework to show that the combination of both the constraints and the training of the adjacency matrix leads to optimal performances. [ABSTRACT FROM AUTHOR]

Knowledge graphs exhibit a typical hierarchical structure and find extensive applications in various artificial intelligence domains. However, large-scale knowledge graphs need to be completed, which limits the performance of knowledge graphs in downstream tasks. Knowledge graph embedding methods have emerged as a primary solution to enhance knowledge graph completeness. These methods aim to represent entities and relations as low-dimensional vectors, focusing on handling relation patterns and multi-relation types. Researchers need to pay more attention to the crucial feature of hierarchical relationships in real-world knowledge graphs. We propose a novel knowledge graph embedding model called Hierarchy-Aware Paired Relation Vectors Knowledge Graph Embedding (HPRE) to bridge this gap. By leveraging the power of 2D coordinates, HPRE adeptly model relation patterns, multi-relation types, and hierarchical features in the knowledge graph. Specifically, HPRE employs paired relation vectors to capture the distinct characteristics of head and tail entities, facilitating a better fit for relational patterns and multi-relation scenarios. Additionally, HPRE employs angular coordinates to differentiate entities at various levels of the hierarchy, effectively representing the hierarchical nature of the knowledge graph. The experimental results show that the HPRE model can effectively learn the hierarchical features of the knowledge graph and achieve state-of-the-art experimental results on multiple real-world datasets for the link prediction task. [ABSTRACT FROM AUTHOR]

Link prediction remains paramount in knowledge graph embedding (KGE), aiming to discern obscured or non-manifest relationships within a given knowledge graph (KG). Despite the critical nature of this endeavor, contemporary methodologies grapple with notable constraints, predominantly in terms of computational overhead and the intricacy of encapsulating multifaceted relationships. This paper introduces a sophisticated approach that amalgamates convolutional operators with pertinent graph structural information. By meticulously integrating information pertinent to entities and their immediate relational neighbors, we enhance the performance of the convolutional model, culminating in an averaged embedding ensuing from the convolution across entities and their proximal nodes. Significantly, our methodology presents a distinctive avenue, facilitating the inclusion of edge-specific data into the convolutional model's input, thus endowing users with the latitude to calibrate the model's architecture and parameters congruent with their specific dataset. Empirical evaluations underscore the ascendancy of our proposition over extant convolution-based link prediction benchmarks, particularly evident across the FB15k, WN18, and YAGO3-10 datasets. The primary objective of this research lies in forging KGE link prediction methodologies imbued with heightened efficiency and adeptness, thereby addressing salient challenges inherent to real-world applications. [ABSTRACT FROM AUTHOR]

Link prediction is an important problem in complex network analysis, which can discover missing or possible future edges in the network. In recent years, link prediction methods based on network representation learning have made progress. But there are two problems with these methods. One is that neighborhood-based node sampling methods cannot handle the situation between two nodes that do not have any common neighbors. The other is the Skip-Gram-based embedding model that represents nodes as static vectors, which cannot reflect the various meanings of nodes. To overcome these two limitations, this paper proposes a method called AFS-BERT (Information entropy based Adaptive Fusion Sampling and BERT embedding model). First, this method defines a centrality score based on adjacency information entropy, which reflects the global and local importance of nodes. Second, we propose a sampling method that adaptively fuses two different strategies using the centrality score. Finally, the BERT-based embedding model is used to realize the low-dimensional dynamic vector representation of nodes. Experimental result on six real-world network datasets shows that AFS-BERT has better performance. Compared with methods of the same type, AFS-BERT achieves upto 6.7% improvement. [ABSTRACT FROM AUTHOR]

Drug–drug interaction (DDI) prediction is one of the essential tasks in drug development to ensure public health and patient safety. Drug combinations with potentially severe DDIs have been verified to threaten the safety of patients critically, and it is therefore of great significance to develop effective computational algorithms for identifying potential DDIs in clinical trials. By modeling DDIs with a graph structure, recent attempts have been made to solve the prediction problem of DDIs by using advanced graph representation learning techniques. Still, their representational capacity is limited by isomorphic structures that are frequently observed in DDI networks. To address this problem, we propose a novel algorithm called DDIGIN to predict DDIs by incorporating a graph isomorphism network (GIN) such that more discriminative representations of drugs can thus be learned for improved performance. Given a DDI network, DDIGIN first initializes the representations of drugs with Node2Vec according to the topological structure and then optimizes these representations by propagating and aggregating the first-order neighboring information in an injective way. By doing so, more powerful representations can thus be learned for drugs with isomorphic structures. Last, DDIGIN estimates the interaction probability for pairwise drugs by multiplying their representations in an end-to-end manner. Experimental results demonstrate that DDIGIN outperforms several state-of-the-art algorithms on the ogbl-ddi (Acc = 0.8518, AUC = 0.8594, and AUPR = 0.9402) and DDInter datasets (Acc = 0.9763, AUC = 0.9772, and AUPR = 0.9868). In addition, our case study indicates that incorporating GIN enhances the expressive power of drug representations for improved performance of DDI prediction. [ABSTRACT FROM AUTHOR]

Many complex systems such as the Internet, the World Wide Web, the brain, and the causes of diseases can be described by networks with nodes representing agents and links representing relationships or interactions between nodes. Despite these systems seeming utterly different at first glance, they are all made up of interacting parts. Individual objects in this type of system are not isolated but connected through links or relationships. Long non-coding RNAs (LncRNAs), one of the factors related to many diseases, are specific genes in the human genome that control many different biological processes. lncRNAs have been shown to regulate cancer development and occurrence. We used link prediction bioinformatics tools to identify lncRNAs affecting various types of cancers. To achieve this goal, two-part networks were used using CN (Common Neighbors), AA (Adamic/Adar), PA (Preferential Attachment), and JC (Jaccard's Coefficient) algorithms. The results indicated that all the obtained lncRNAs with a high score had been reviewed in other research articles, which was a sign of the correctness of our algorithms. 4.5% of the obtained results lacked scientific and research studies, all with a high score for future studies and included lncRNA H19 and SPRY4-IT1. In addition, in one case with a high score and first position in the Excel file obtained using the Jaccard coefficient algorithm, lncRNA with the symbol AC09510.3 was associated with adenocarcinoma. Still, this possible link has not been investigated in any article yet. Conclusion: The current study may identify novel lncRNAs implicated in Adenocarcinoma, vulva squamous cell carcinoma, basal cell carcinoma, gestational choriocarcinoma, chromophobe renal cell carcinoma, brain cancer, pancreatic cancer, hepatocellular carcinoma, prostate cancer, embryonal cancer, germ cell cancer, and choriocarcinoma; however, further research is required to determine the potential functions of this lncRNAs in cancers mentioned above. [ABSTRACT FROM AUTHOR]

Drug–drug interaction (DDI) identification is essential to clinical medicine and drug discovery. The two categories of drugs (i.e. chemical drugs and biotech drugs) differ remarkably in molecular properties, action mechanisms, etc. Biotech drugs are up-to-comers but highly promising in modern medicine due to higher specificity and fewer side effects. However, existing DDI prediction methods only consider chemical drugs of small molecules, not biotech drugs of large molecules. Here, we build a large-scale dual-modal graph database named CB-DB and customize a graph-based framework named CB-TIP to reason event-aware DDIs for both chemical and biotech drugs. CB-DB comprehensively integrates various interaction events and two heterogeneous kinds of molecular structures. It imports endogenous proteins founded on the fact that most drugs take effects by interacting with endogenous proteins. In the modality of molecular structure, drugs and endogenous proteins are two heterogeneous kinds of graphs, while in the modality of interaction, they are nodes connected by events (i.e. edges of different relationships). CB-TIP employs graph representation learning methods to generate drug representations from either modality and then contrastively mixes them to predict how likely an event occurs when a drug meets another in an end-to-end manner. Experiments demonstrate CB-TIP 's great superiority in DDI prediction and the promising potential of uncovering novel DDIs. [ABSTRACT FROM AUTHOR]

Link prediction (LP) is an attractive research problem on social network data. Yet, the link information may be leaked by the untrusted collector. As countermeasures, there are a few methods specially designed for link prediction under local differential privacy (LDP), which allow the third-party data collector collects user data for link prediction while protecting user connection privacy. In this paper, we propose a C ommunity-based graph collection with P ersonalized sampling R andomized R esponse (CPRR) which is a novel graph collection algorithm with LDP and reaches decent trade-off between sensitive link protection and link prediction performance. The proposed mechanism adopts a personalized sampling technique for each connection of a user and then utlizes the randomized response mechanism on the sampled subset. Based on the personalized sampling technique, we can reduce the injected noise in LDP. Meanwhile, considering the different edge distributions of the different regions on the original graph, we propose a community-based sampling strategy. Then, we prove that the proposed CPRR satisfies LDP. Through extensive experiments on several real-life graph datasets, we demonstrate that CPRR can achieve better results in balancing privacy protection and link prediction performance than the state-of-art baselines. [ABSTRACT FROM AUTHOR]