Network Embedding Research Articles

Product Matching with Two-Branch Neural Network Embedding

Product Matching with Two-Branch Neural Network Embedding

Multi-Scale Spatio-Temporal Attention Networks for Network-Scale Traffic Learning and Forecasting.

Accurate and timely forecasting of traffic on local road networks is crucial for deploying effective dynamic traffic control, advanced route planning, and navigation services. This task is particularly challenging due to complex spatio-temporal dependencies arising from non-Euclidean spatial relations in road networks and non-linear temporal dynamics influenced by changing road conditions. This paper introduces the spatio-temporal network embedding (STNE) model, a novel deep learning framework tailored for learning and forecasting graph-structured traffic data over extended input sequences. Unlike traditional convolutional neural networks (CNNs), the model employs graph convolutional networks (GCNs) to capture the spatial characteristics of local road network topologies. Moreover, the segmentation of very long input traffic data into multiple sub-sequences, based on significant temporal properties such as closeness, periodicity, and trend, is performed. Multi-dimensional long short-term memory neural networks (MDLSTM) are utilized to flexibly access multi-dimensional context. Experimental results demonstrate that the STNE model surpasses state-of-the-art traffic forecasting benchmarks on two large-scale real-world traffic datasets.

IMI-driver: Integrating multi-level gene networks and multi-omics for cancer driver gene identification.

The identification of cancer driver genes is crucial for early detection, effective therapy, and precision medicine of cancer. Cancer is caused by the dysregulation of several genes at various levels of regulation. However, current techniques only capture a limited amount of regulatory information, which may hinder their efficacy. In this study, we present IMI-driver, a model that integrates multi-omics data into eight biological networks and applies Multi-view Collaborative Network Embedding to embed the gene regulation information from the biological networks into a low-dimensional vector space to identify cancer drivers. We apply IMI-driver to 29 cancer types from The Cancer Genome Atlas (TCGA) and compare its performance with nine other methods on nine benchmark datasets. IMI-driver outperforms the other methods, demonstrating that multi-level network integration enhances prediction accuracy. We also perform a pan-cancer analysis using the genes identified by IMI-driver, which confirms almost all our selected candidate genes as known or potential drivers. Case studies of the new positive genes suggest their roles in cancer development and progression.

Neighbor-Enhanced Representation Learning for Link Prediction in Dynamic Heterogeneous Attributed Networks

Dynamic link prediction aims to predict future connections among unconnected nodes in a network. It can be applied for friend recommendations, link completion, and other tasks. Network representation learning algorithms have demonstrated considerable effectiveness in various prediction tasks. However, most network representation learning algorithms are based on homogeneous networks and static networks for link prediction that do not consider rich semantic and dynamic information. Additionally, existing dynamic network representation learning methods neglect the neighborhood interaction structure of the node. In this work, we design a neighbor-enhanced dynamic heterogeneous attributed network embedding method (NeiDyHNE) for link prediction. In light of the impressive achievements of the heuristic methods, we learn the information of common neighbors and neighbors’ interaction in heterogeneous networks to preserve the neighbors proximity and common neighbors proximity. NeiDyHNE encodes the attributes and neighborhood structure of nodes as well as the evolutionary features of the dynamic network. More specifically, NeiDyHNE consists of the hierarchical structure attention module and the convolutional temporal attention module. The hierarchical structure attention module captures the rich features and semantic structure of nodes. The convolutional temporal attention module captures the evolutionary features of the network over time in dynamic heterogeneous networks. We evaluate our method and various baseline methods on the dynamic link prediction task. Experimental results demonstrate that our method is superior to baseline methods in terms of accuracy.

AGENDA: Predicting Trip Purposes with A New Graph Embedding Network and Active Domain Adaptation

Trip purpose is a meaningful aspect of travel behaviour for the understanding of urban mobility. However, it is non-trivial to automatically obtain trip purposes. On one hand, trip purposes are naturally diverse and complicated, but the available predictive data sources are limited in real-world scenarios. On the other hand, since trip purpose labeling is costly and the development levels of cities are unbalanced, it is infeasible to access large-scale labeled data in less developed cities to train advanced prediction models. To narrow the gaps, this article presents A new Graph Embedding Network and active Domain Adaptation based framework (AGENDA) that only requires open data sources and is capable of predicting in both label-rich cities and label-scarce cities. Specifically, in label-rich source cities, we first use the vehicle’s GPS trajectory and open POI check-ins to augment trip contexts. Then we establish a supervised graph embedding network with two attention mechanisms to extract the passenger’s latent activity semantics and a classifier to predict trip purpose. To enable the prediction in label-scarce target cities, we further devise an active domain adaptation framework, in which adversarial domain adaptation is used to transfer the source-learned knowledge, and active learning is used to integrate human intelligence in the model training. A group of experiments are conducted with real-world datasets in Beijing and Shanghai. Evaluation results demonstrate that the proposed framework significantly outperforms existing trip purpose prediction algorithms, and could make accurate trip purpose prediction in label-scarce cities with much fewer labeling efforts.

Multimodal Recipe Recommendation with Heterogeneous Graph Neural Networks

Recipe recommendation is the process of recommending suitable recipes to users based on factors such as user preferences and dietary needs. Recipes typically involve multiple modalities, with text and images being common, while most typical recipe recommendation methods recommend recipes to users based on text. Obviously, the expressiveness of a single modal is often not enough, and the semantic information of images is more abundant. Moreover, it is difficult to grasp the feature fusion granularity of different kinds of modal information and the relationship between recipes and users. To solve the above problem, this paper proposes a Multimodal Heterogeneous Graph Neural Network Recipe Recommendation (MHGRR) architecture, which aims to fully fuse the various kinds of modal information of recipes and handle the relationship between users and recipes. We use embedding and shallow Convolutional Neural Networks (CNNs) to extract original text and image information for unifying feature fusion granularity, and use Heterogeneous Graph Neural Networks based on GraphSAGE to capture the complex relationship between users and recipes. To verify the effectiveness of our proposed model, we perform some comparative experiments on a real dataset; the experiments show that our method outperforms most popular recipe recommendation methods. Through an ablation experiment, we found that adding image information to recipe recommendation is more effective, and we additionally found that as the output dimensions of GraphSAGE increased, the performance of the model varied little.

DeepAGS: Deep learning with activity, geography and sequential information in predicting an individual's next trip destination

AbstractIndividual mobility is driven by activities and thus restricted geographically, especially for trip destination prediction in public transport. Existing statistical learning based models focus on extracting mobility regularity in predicting an individual's mobility. However, they are limited in modeling varied spatial mobility patterns driven by the same activity (e.g. an individual may travel to different locations for shopping). The paper proposes a deep learning model with activity, geographic and sequential (DeepAGS) information in predicting an individual's next trip destination in public transport. DeepAGS models the semantic features of activity and geography by using word embedding and graph convolutional network. An adaptive neural fusion gate mechanism is proposed to dynamically fuse the mobility activity and geographical information given the current trip information. Besides, DeepAGS uses the gated recurrent unit to capture the temporal mobility regularity. The approach is validated by using a real‐world smartcard dataset in urban railway systems and comparing with state‐of‐the‐art models. The results show that the proposed model outperforms its peers in terms of accuracy and robustness by effectively integrating the activity and geographical information relevant to a trip context. Also, we illustrate and verify the working mechanism of the DeepAGS model using the synthetic data constructed using real‐world data. The DeepAGS model captures both the activity and geographic information of hidden mobility activities and thus could be potentially applicable to other mobility prediction tasks, such as bus trip destinations and individual GPS locations.

Analyzing public sentiment toward economic stimulus using natural language processing

PurposeThis study aims to investigate public sentiment toward economic stimulus using textual analysis. Specifically, it analyzes Twitter’s public opinion, emotion-based sentiment and topics related to COVID-19 economic stimulus packages.Design/methodology/approachThis study applies natural language processing techniques, such as sentiment analysis, t-distributed stochastic neighbor embedding and semantic network analysis, to a global data set of 88,441 tweets from January 2020 to December 2021 extracted from the Twitter platform, discussing COVID-19 economic stimulus packages.FindingsResults show that in the fourth quarter of 2021, there is a declining trend of positive sentiment (−5%) and an increasing trend of negative sentiment (14%), which may indicate the perceived inadequacy of COVID-19 stimulus measures. Topic modeling identifies seven topics, highlighting the necessity of stimulus in the education sector.Practical implicationsThe big-data findings of this study provide a better understanding of public sentiment about economic stimulus for regulators and policymakers, which can help in formulating more effective fiscal and monetary policies.Originality/valuePublic sentiment is a significant concern for regulators because of its associated ambiguity, such as how to design stimulus packages and evaluate the effectiveness of previous measures. This study applies natural language processing, contributing to the growing literature on designing effective economic stimulus.

XG and CS-based self-assembled nanocomposite hydrogel embedding fluorescent NCQDs capable of detection and adsorptive removal of the polar MO and Cr(VI) pollutants

Aiming at dealing with organic and inorganic pollutants dissolved in aquatic environments, we introduce self-assembled fluorescent nanocomposite hydrogel based on a binary polysaccharide network (xanthan gum/chitosan) embedding nitrogen-doped carbon quantum dots not only as a hybrid solid optical sensor for detecting Cr(VI) ions but also to remove anionically charged contaminants Cr(VI) and methyl orange (MO) by acting as an adsorbent. This fluorescent nanocomposite achieved a detection limit of 0.29 μM when used to detect Cr(VI) and demonstrated a fluorescence quantum yield of 59.7 %. Several factors contributed to the effectiveness of the adsorption of Cr(VI) and MO in batch studies, including the solution pH, dosage of the adsorbent, temperature, initial contamination level, and contact time. Experimental results showed 456 mg/g maximum adsorption capacity at pH 4 for MO compared to 291 mg/g at pH 2 for Cr(VI) at 25 °C. In addition to conforming to Langmuir's model, Cr(VI) and MO's adsorption kinetics closely matched pseudo-second-order. Using thermodynamic parameters, the results indicate that Cr(VI) and MO adsorb spontaneously and exothermically. Recycling spent adsorbent for Cr(VI) and MO using NaOH at 0.1 M was possible; the respective adsorption efficiency remained at approximately 82.2 % and 83 % after the fifth regeneration cycle.

Energy efficient resource allocation based on virtual network embedding for IoT data generation

Energy efficient resource allocation based on virtual network embedding for IoT data generation

HGEE: Learning for Trajectory Prediction with Heterogeneous Graph Interaction and External Embedding of Unmanned Swarm Systems in Adversarial Environment

Trajectory prediction of unmanned swarm systems, serving as the foundation for behavioral and intentional cognition, has attracted extensive attention and made considerable progress in adversarial research. The influence of heterogeneous interaction relationships and external factors is crucial for trajectory prediction. Consequently, this paper proposes the Heterogeneous Graph with External Embedding (HGEE) network. We model the latent variables as multi-layer heterogeneous graphs based on prior knowledge of different interaction relationships and propose a method for calculating edge embeddings for heterogeneous graphs. Furthermore, we introduce a method that combines external environmental feature with historical observational trajectory data as the input for the decoder, enabling the model to learn the impacts of obstacles, targets, and desired formations on trajectories. We demonstrate that our approach surpasses state-of-the-art models in interaction inference and trajectory prediction through experiments on our proposed formation datasets based on consensus theory, across five evaluation metrics.

Improving the evidential value of low-quality face images with aggregation of deep neural network embeddings

In forensic facial comparison, questioned-source images are usually captured in uncontrolled environments, with non-uniform lighting, and from non-cooperative subjects. The poor quality of such material usually compromises their value as evidence in legal proceedings. On the other hand, in forensic casework, multiple images of the person of interest are usually available. In this paper, we propose to aggregate deep neural network embeddings from various images of the same person to improve the performance in forensic comparison of facial images. We observe significant performance improvements, especially for low-quality images. Further improvements are obtained by aggregating embeddings of more images and by applying quality-weighted aggregation. We demonstrate the benefits of this approach in forensic evaluation settings with the development and validation of common-source likelihood ratio systems and report improvements in Cllr both for CCTV images and for social media images.

Z-Laplacian Matrix Factorization: Network Embedding With Interpretable Graph Signals

Z-Laplacian Matrix Factorization: Network Embedding With Interpretable Graph Signals

Scalable Solution for Service Function Chaining Embedding in Hybrid Optical-Electronic Networks

Scalable Solution for Service Function Chaining Embedding in Hybrid Optical-Electronic Networks

Kilonova Spectral Inverse Modelling with Simulation-based Inference: An Amortized Neural Posterior Estimation Analysis

Kilonovae represent a category of astrophysical transients, identifiable as the electromagnetic (EM) counterparts associated with the coalescence events of binary systems comprising neutron stars and neutron star–black hole pairs. They act as probes for heavy-element nucleosynthesis in astrophysical environments. These studies rely on an inference of the physical parameters (e.g., ejecta mass, velocity, composition) that describe kilonovae-based on EM observations. This is a complex inverse problem typically addressed with sampling-based methods such as Markov Chain Monte Carlo or nested sampling algorithms. However, repeated inferences can be computationally expensive, due to the sequential nature of these methods. This poses a significant challenge to ensuring the reliability and statistical validity of the posterior approximations and, thus, the inferred kilonova parameters themselves. We present a novel approach: simulation-based inference using simulations produced by KilonovaNet. Our method employs an ensemble of amortized neural posterior estimation (ANPE) with an embedding network to directly predict posterior distributions from simulated spectral energy distributions. We take advantage of the quasi-instantaneous inference time of ANPE to demonstrate the reliability of our posterior approximations using diagnostics tools, including coverage diagnostic and posterior predictive checks. We further test our model with real observations from AT 2017gfo, the only kilonova with multimessenger data, demonstrating agreement with previous likelihood-based methods while reducing inference time down to a few seconds. The inference results produced by ANPE appear to be conservative and reliable, paving the way for testable and more efficient kilonova parameter inference.

Research on group decision optimization and influence analysis in social network based on machine learning

With the rapid development of online social networks, the research on group decision-making in social networks has attracted extensive attention. Social networks facilitate interaction and behavior between individuals, businesses, and organizations. However, traditional group decision-making methods often ignore the social relationships between group members and fail to fully consider the impact of these relationships on subgroup division. In this work, we propose a novel model approach that combines graph neural networks and deep learning techniques to capture and analyze complex relational structures in social networks. The model uses node features and edge features to optimize the group decision-making process and effectively evaluate the influence between individuals through multi-layer network embedding and aggregation operations. Experimental analysis results show that the proposed method performs well in improving the accuracy and efficiency of decision-making, and significantly improves the quality of decision-making.

Iterative embedding and reweighting of complex networks reveals community structure

Graph embeddings learn the structure of networks and represent it in low-dimensional vector spaces. Community structure is one of the features that are recognized and reproduced by embeddings. We show that an iterative procedure, in which a graph is repeatedly embedded and its links are reweighted based on the geometric proximity between the nodes, reinforces intra-community links and weakens inter-community links, making the clusters of the initial network more visible and more easily detectable. The geometric separation between the communities can become so strong that even a very simple parsing of the links may recover the communities as isolated components with surprisingly high precision. Furthermore, when used as a pre-processing step, our embedding and reweighting procedure can improve the performance of traditional community detection algorithms.

HTINet2: herb-target prediction via knowledge graph embedding and residual-like graph neural network.

Target identification is one of the crucial tasks in drug research and development, as it aids in uncovering the action mechanism of herbs/drugs and discovering new therapeutic targets. Although multiple algorithms of herb target prediction have been proposed, due to the incompleteness of clinical knowledge and the limitation of unsupervised models, accurate identification for herb targets still faces huge challenges of data and models. To address this, we proposed a deep learning-based target prediction framework termed HTINet2, which designed three key modules, namely, traditional Chinese medicine (TCM) and clinical knowledge graph embedding, residual graph representation learning, and supervised target prediction. In the first module, we constructed a large-scale knowledge graph that covers the TCM properties and clinical treatment knowledge of herbs, and designed a component of deep knowledge embedding to learn the deep knowledge embedding of herbs and targets. In the remaining two modules, we designed a residual-like graph convolution network to capture the deep interactions among herbs and targets, and a Bayesian personalized ranking loss to conduct supervised training and target prediction. Finally, we designed comprehensive experiments, of which comparison with baselines indicated the excellent performance of HTINet2 (HR@10 increased by 122.7% and NDCG@10 by 35.7%), ablation experiments illustrated the positive effect of our designed modules of HTINet2, and case study demonstrated the reliability of the predicted targets of Artemisia annua and Coptis chinensis based on the knowledge base, literature, and molecular docking.

DyHDGE: Dynamic heterogeneous transaction graph embedding for safety-centric fraud detection in financial scenarios

Dynamic graph fraud detection aims to distinguish fraudulent entities that deviate significantly from most benign entities within an ever-changing graph network. However, when dealing with different financial fraud scenarios, existing methods face challenges, resulting in difficulty in effectively ensuring financial security. In fraud scenarios, transaction data are generated in real time, in which a strong temporal relationship between multiple fraudulent transactions is observed. Traditional dynamic graph models struggle to effectively balance the temporal features of nodes and spatial structural features, failing to handle different types of nodes in the graph network. In this study, to extract the temporal and structural information, we proposed a dynamic heterogeneous transaction graph embedding (DyHDGE) network based on a dynamic heterogeneous transaction graph, considering both temporal and structural information while incorporating heterogeneous data. To separately extract temporal relationships between transactions and spatial structural relationships between nodes, we used a heterogeneous temporal graph representation learning module and a temporal graph structure information extraction module. Additionally, we designed two loss functions to optimize node feature representations. Extensive experiments demonstrated that the proposed DyHDGE significantly outperformed previous state-of-the-art methods on two simulated datasets of financial fraud scenarios. This capability contributes to enhancing security in financial consumption scenarios.

MENTOR: Multiplex Embedding of Networks for Team-Based Omics Research.

While the proliferation of data-driven omics technologies has continued to accelerate, methods of identifying relationships among large-scale changes from omics experiments have stagnated. It is therefore imperative to develop methods that can identify key mechanisms among one or more omics experiments in order to advance biological discovery. To solve this problem, here we describe the network-based algorithm MENTOR - Multiplex Embedding of Networks for Team-Based Omics Research. We demonstrate MENTOR's utility as a supervised learning approach to successfully partition a gene set containing multiple ontological functions into their respective functions. Subsequently, we used MENTOR as an unsupervised learning approach to identify important biological functions pertaining to the host genetic architectures in Populus trichocarpa associated with microbial abundance of multiple taxa. Moreover, as open source software designed with scientific teams in mind, we demonstrate the ability to use the output of MENTOR to facilitate distributed interpretation of omics experiments.