IDS-INT:Intrusion detection system using transformer-based transfer learning for imbalanced network traffic

A network intrusion detection system is critical for cyber security against llegitimate attacks.In terms of feature perspectives,network traffic may include a variety of elements such as attack reference,attack type,a subcategory of attack,host information,malicious scripts,etc.In terms of network perspectives,network traffic may contain an imbalanced number of harmful attacks when compared to normal traffic.It is challenging to identify a specific attack due to complex features and data imbalance issues.To address these issues,this paper proposes an Intrusion Detection System using transformer-based transfer learning for Imbalanced Network Traffic(IDS-INT).IDS-INT uses transformer-based transfer learning to learn feature interactions in both network feature representation and imbalanced data.First,detailed information about each type of attack is gathered from network interaction descriptions,which include network nodes,attack type,reference,host information,etc.Second,the transformer-based transfer learning approach is developed to learn detailed feature representation using their semantic anchors.Third,the Synthetic Minority Oversampling Technique(SMOTE)is implemented to balance abnormal traffic and detect minority attacks.Fourth,the Convolution Neural Network(CNN)model is designed to extract deep features from the balanced network traffic.Finally,the hybrid approach of the CNN-Long Short-Term Memory(CNN-LSTM)model is developed to detect different types of attacks from the deep features.Detailed experiments are conducted to test the proposed approach using three standard datasets,i.e.,UNsWNB15,CIC-IDS2017,and NSL-KDD.An explainable AI approach is implemented to interpret the proposed method and develop a trustable model.

Efficacy of hepatic arterial infusion chemotherapy and its combination strategies for advanced hepatocellular carcinoma:A network meta-analysis

BACKGROUND With the rapid progress of systematic therapy for hepatocellular carcinoma(HCC),therapeutic strategies combining hepatic arterial infusion chemotherapy(HAIC)with systematic therapy arised increasing concentrations.However,there have been no systematic review comparing HAIC and its combination strategies in the first-line treatment for advanced HCC.AIM To investigate the efficacy and safety of HAIC and its combination therapies for advanced HCC.METHODS A network meta-analysis was performed by including 9 randomized controlled trails and 35 cohort studies to carry out our study.The outcomes of interest comprised overall survival(OS),progression-free survival(PFS),tumor response and adverse events.Hazard ratios(HR)and odds ratios(OR)with a 95% confidence interval(CI)were calculated and agents were ranked based on their ranking probability.RESULTS HAIC outperformed Sorafenib(HR=0.55,95%CI:0.42-0.72;HR=0.51,95%CI:0.33-0.78;OR=2.86,95%CI:1.37-5.98;OR=5.45,95%CI:3.57-8.30;OR=7.15,95%CI:4.06-12.58;OR=2.89,95%CI:1.99-4.19;OR=0.48,95%CI:0.25-0.92,respectively)and transarterial chemoembolization(TACE)(HR=0.50,95%CI:0.33-0.75;HR=0.62,95%CI:0.39-0.98;OR=3.08,95%CI:1.36-6.98;OR=2.07,95%CI:1.54-2.80;OR=3.16,95%CI:1.71-5.85;OR=2.67,95%CI:1.59-4.50;OR=0.16,95%CI:0.05-0.54,respectively)in terms of efficacy and safety.HAIC+lenvatinib+ablation,HAIC+ablation,HAIC+anti-programmed cell death 1(PD-1),and HAIC+radiotherapy had the higher likelihood of providing better OS and PFS outcomes compared to HAIC alone.HAIC+TACE+S-1,HAIC+lenvatinib,HAIC+PD-1,HAIC+TACE,and HAIC+sorafenib had the higher likelihood of providing better partial response and objective response rate outcomes compared to HAIC.HAIC+PD-1,HAIC+TACE+S-1 and HAIC+TACE had the higher likelihood of providing better complete response and disease control rate outcomes compared to HAIC alone.CONCLUSION HAIC proved more effective and safer than sorafenib and TACE.Furthermore,combined with other interventions,HAIC showed improved efficacy over HAIC monotherapy according to the treatment ranking analysis.

Multi-scale physics-informed neural networks for solving high Reynolds number boundary layer flows based on matched asymptotic expansions

Multi-scale system remains a classical scientific problem in fluid dynamics,biology,etc.In the present study,a scheme of multi-scale Physics-informed neural networks is proposed to solve the boundary layer flow at high Reynolds numbers without any data.The flow is divided into several regions with different scales based on Prandtl's boundary theory.Different regions are solved with governing equations in different scales.The method of matched asymptotic expansions is used to make the flow field continuously.A flow on a semi infinite flat plate at a high Reynolds number is considered a multi-scale problem because the boundary layer scale is much smaller than the outer flow scale.The results are compared with the reference numerical solutions,which show that the msPINNs can solve the multi-scale problem of the boundary layer in high Reynolds number flows.This scheme can be developed for more multi-scale problems in the future.

Non-crossing Quantile Regression Neural Network as a Calibration Tool for Ensemble Weather Forecasts

Despite the maturity of ensemble numerical weather prediction(NWP),the resulting forecasts are still,more often than not,under-dispersed.As such,forecast calibration tools have become popular.Among those tools,quantile regression(QR)is highly competitive in terms of both flexibility and predictive performance.Nevertheless,a long-standing problem of QR is quantile crossing,which greatly limits the interpretability of QR-calibrated forecasts.On this point,this study proposes a non-crossing quantile regression neural network(NCQRNN),for calibrating ensemble NWP forecasts into a set of reliable quantile forecasts without crossing.The overarching design principle of NCQRNN is to add on top of the conventional QRNN structure another hidden layer,which imposes a non-decreasing mapping between the combined output from nodes of the last hidden layer to the nodes of the output layer,through a triangular weight matrix with positive entries.The empirical part of the work considers a solar irradiance case study,in which four years of ensemble irradiance forecasts at seven locations,issued by the European Centre for Medium-Range Weather Forecasts,are calibrated via NCQRNN,as well as via an eclectic mix of benchmarking models,ranging from the naïve climatology to the state-of-the-art deep-learning and other non-crossing models.Formal and stringent forecast verification suggests that the forecasts post-processed via NCQRNN attain the maximum sharpness subject to calibration,amongst all competitors.Furthermore,the proposed conception to resolve quantile crossing is remarkably simple yet general,and thus has broad applicability as it can be integrated with many shallow-and deep-learning-based neural networks.

Attention-based network embedding with higher-order weights and node attributes

Network embedding aspires to learn a low-dimensional vector of each node in networks,which can apply to diverse data mining tasks.In real-life,many networks include rich attributes and temporal information.However,most existing embedding approaches ignore either temporal information or network attributes.A self-attention based architecture using higher-order weights and node attributes for both static and temporal attributed network embedding is presented in this article.A random walk sampling algorithm based on higher-order weights and node attributes to capture network topological features is presented.For static attributed networks,the algorithm incorporates first-order to k-order weights,and node attribute similarities into one weighted graph to preserve topological features of networks.For temporal attribute networks,the algorithm incorporates previous snapshots of networks containing first-order to k-order weights,and nodes attribute similarities into one weighted graph.In addition,the algorithm utilises a damping factor to ensure that the more recent snapshots allocate a greater weight.Attribute features are then incorporated into topological features.Next,the authors adopt the most advanced architecture,Self-Attention Networks,to learn node representations.Experimental results on node classification of static attributed networks and link prediction of temporal attributed networks reveal that our proposed approach is competitive against diverse state-of-the-art baseline approaches.

Multi-layer network embedding on scc-based network with motif

Interconnection of all things challenges the traditional communication methods,and Semantic Communication and Computing(SCC)will become new solutions.It is a challenging task to accurately detect,extract,and represent semantic information in the research of SCC-based networks.In previous research,researchers usually use convolution to extract the feature information of a graph and perform the corresponding task of node classification.However,the content of semantic information is quite complex.Although graph convolutional neural networks provide an effective solution for node classification tasks,due to their limitations in representing multiple relational patterns and not recognizing and analyzing higher-order local structures,the extracted feature information is subject to varying degrees of loss.Therefore,this paper extends from a single-layer topology network to a multi-layer heterogeneous topology network.The Bidirectional Encoder Representations from Transformers(BERT)training word vector is introduced to extract the semantic features in the network,and the existing graph neural network is improved by combining the higher-order local feature module of the network model representation network.A multi-layer network embedding algorithm on SCC-based networks with motifs is proposed to complete the task of end-to-end node classification.We verify the effectiveness of the algorithm on a real multi-layer heterogeneous network.

Intrusion Detection System for Smart Industrial Environments with Ensemble Feature Selection and Deep Convolutional Neural Networks

Smart Industrial environments use the Industrial Internet of Things(IIoT)for their routine operations and transform their industrial operations with intelligent and driven approaches.However,IIoT devices are vulnerable to cyber threats and exploits due to their connectivity with the internet.Traditional signature-based IDS are effective in detecting known attacks,but they are unable to detect unknown emerging attacks.Therefore,there is the need for an IDS which can learn from data and detect new threats.Ensemble Machine Learning(ML)and individual Deep Learning(DL)based IDS have been developed,and these individual models achieved low accuracy;however,their performance can be improved with the ensemble stacking technique.In this paper,we have proposed a Deep Stacked Neural Network(DSNN)based IDS,which consists of two stacked Convolutional Neural Network(CNN)models as base learners and Extreme Gradient Boosting(XGB)as the meta learner.The proposed DSNN model was trained and evaluated with the next-generation dataset,TON_IoT.Several pre-processing techniques were applied to prepare a dataset for the model,including ensemble feature selection and the SMOTE technique.Accuracy,precision,recall,F1-score,and false positive rates were used to evaluate the performance of the proposed ensemble model.Our experimental results showed that the accuracy for binary classification is 99.61%,which is better than in the baseline individual DL and ML models.In addition,the model proposed for IDS has been compared with similar models.The proposed DSNN achieved better performance metrics than the other models.The proposed DSNN model will be used to develop enhanced IDS for threat mitigation in smart industrial environments.

A Novel Locomotion Rule Rmbedding Long Short-Term Memory Network with Attention for Human Locomotor Intent Classification Using Multi-Sensors Signals

Locomotor intent classification has become a research hotspot due to its importance to the development of assistive robotics and wearable devices.Previous work have achieved impressive performance in classifying steady locomotion states.However,it remains challenging for these methods to attain high accuracy when facing transitions between steady locomotion states.Due to the similarities between the information of the transitions and their adjacent steady states.Furthermore,most of these methods rely solely on data and overlook the objective laws between physical activities,resulting in lower accuracy,particularly when encountering complex locomotion modes such as transitions.To address the existing deficiencies,we propose the locomotion rule embedding long short-term memory(LSTM)network with Attention(LREAL)for human locomotor intent classification,with a particular focus on transitions,using data from fewer sensors(two inertial measurement units and four goniometers).The LREAL network consists of two levels:One responsible for distinguishing between steady states and transitions,and the other for the accurate identification of locomotor intent.Each classifier in these levels is composed of multiple-LSTM layers and an attention mechanism.To introduce real-world motion rules and apply constraints to the network,a prior knowledge was added to the network via a rule-modulating block.The method was tested on the ENABL3S dataset,which contains continuous locomotion date for seven steady and twelve transitions states.Experimental results showed that the LREAL network could recognize locomotor intents with an average accuracy of 99.03%and 96.52%for the steady and transitions states,respectively.It is worth noting that the LREAL network accuracy for transition-state recognition improved by 0.18%compared to other state-of-the-art network,while using data from fewer sensors.

Anion exchange membranes with a semi-interpenetrating polymer network using 1,6-dibromohexane as bifunctional crosslinker

An anion exchange membrane(AEM)is generally expected to possess high ion exchange capacity(IEC),low water uptake(WU),and high mechanical strength when applied to electrodialysis desalination.Among different types of AEMs,semi-interpenetrating polymer networks(SIPNs)have been suggested for their structural superiorities,i.e.,the tunable local density of ion exchange groups for IEC and the restrained leaching of hygroscopic groups by insolubility for WU.Unfortunately,the conventional SIPN AEMs still struggle to balances IEC,WU,and mechanical strength simultaneously,due to the lack of the compact crosslinking region.In this work,we proposed a novel SIPN structure of polyvinylidene difluoride/polyvinylimidazole/1,6-dibromohexane(PVDF/PVIm/DBH).On the one hand,DBH with two cationic groups of imidazole groups are introduced to enhance the ion conductivity,which is different from the conventional monofunctional modifier with only one cationic group.On the other hand,DBH has the ability to bridge with PVIm,where the mechanical strength of the resulting AEM is increased by the increase of crosslinking degree.Results show that a low WU of 38.1%to 62.6%,high IEC of 2.12—2.22 mmol·g^(-1),and excellent tensile strength of 3.54—12.35 MPa for PVDF/PVIm/DBH membrane are achieved.This work opens a new avenue for achieving the high-quality AEMs.

Combining reinforcement learning with mathematical programming:An approach for optimal design of heat exchanger networks

Heat integration is important for energy-saving in the process industry.It is linked to the persistently challenging task of optimal design of heat exchanger networks(HEN).Due to the inherent highly nonconvex nonlinear and combinatorial nature of the HEN problem,it is not easy to find solutions of high quality for large-scale problems.The reinforcement learning(RL)method,which learns strategies through ongoing exploration and exploitation,reveals advantages in such area.However,due to the complexity of the HEN design problem,the RL method for HEN should be dedicated and designed.A hybrid strategy combining RL with mathematical programming is proposed to take better advantage of both methods.An insightful state representation of the HEN structure as well as a customized reward function is introduced.A Q-learning algorithm is applied to update the HEN structure using theε-greedy strategy.Better results are obtained from three literature cases of different scales.

Virtual Assembly Collision Detection Algorithm Using Backpropagation Neural Network

As computer graphics technology continues to advance,Collision Detection(CD)has emerged as a critical element in fields such as virtual reality,computer graphics,and interactive simulations.CD is indispensable for ensuring the fidelity of physical interactions and the realism of virtual environments,particularly within complex scenarios like virtual assembly,where both high precision and real-time responsiveness are imperative.Despite ongoing developments,current CD techniques often fall short in meeting these stringent requirements,resulting in inefficiencies and inaccuracies that impede the overall performance of virtual assembly systems.To address these limitations,this study introduces a novel algorithm that leverages the capabilities of a Backpropagation Neural Network(BPNN)to optimize the structural composition of the Hybrid Bounding Volume Tree(HBVT).Through this optimization,the research proposes a refined Hybrid Hierarchical Bounding Box(HHBB)framework,which is specifically designed to enhance the computational efficiency and precision of CD processes.The HHBB framework strategically reduces the complexity of collision detection computations,thereby enabling more rapid and accurate responses to collision events.Extensive experimental validation within virtual assembly environments reveals that the proposed algorithm markedly improves the performance of CD,particularly in handling complex models.The optimized HBVT architecture not only accelerates the speed of collision detection but also significantly diminishes error rates,presenting a robust and scalable solution for real-time applications in intricate virtual systems.These findings suggest that the proposed approach offers a substantial advancement in CD technology,with broad implications for its application in virtual reality,computer graphics,and related fields.

A Fractional-Order Ultra-Local Model-Based Adaptive Neural Network Sliding Mode Control of n-DOF Upper-Limb Exoskeleton With Input Deadzone

This paper proposes an adaptive neural network sliding mode control based on fractional-order ultra-local model for n-DOF upper-limb exoskeleton in presence of uncertainties,external disturbances and input deadzone.Considering the model complexity and input deadzone,a fractional-order ultra-local model is proposed to formulate the original dynamic system for simple controller design.Firstly,the control gain of ultra-local model is considered as a constant.The fractional-order sliding mode technique is designed to stabilize the closed-loop system,while fractional-order time-delay estimation is combined with neural network to estimate the lumped disturbance.Correspondingly,a fractional-order ultra-local model-based neural network sliding mode controller(FO-NNSMC) is proposed.Secondly,to avoid disadvantageous effect of improper gain selection on the control performance,the control gain of ultra-local model is considered as an unknown parameter.Then,the Nussbaum technique is introduced into the FO-NNSMC to deal with the stability problem with unknown gain.Correspondingly,a fractional-order ultra-local model-based adaptive neural network sliding mode controller(FO-ANNSMC) is proposed.Moreover,the stability analysis of the closed-loop system with the proposed method is presented by using the Lyapunov theory.Finally,with the co-simulations on virtual prototype of 7-DOF iReHave upper-limb exoskeleton and experiments on 2-DOF upper-limb exoskeleton,the obtained compared results illustrate the effectiveness and superiority of the proposed method.

Neural Network Robust Control Based on Computed Torque for Lower Limb Exoskeleton

The lower limb exoskeletons are used to assist wearers in various scenarios such as medical and industrial settings.Complex modeling errors of the exoskeleton in different application scenarios pose challenges to the robustness and stability of its control algorithm.The Radial Basis Function(RBF)neural network is used widely to compensate for modeling errors.In order to solve the problem that the current RBF neural network controllers cannot guarantee the asymptotic stability,a neural network robust control algorithm based on computed torque method is proposed in this paper,focusing on trajectory tracking.It innovatively incorporates the robust adaptive term while introducing the RBF neural network term,improving the compensation ability for modeling errors.The stability of the algorithm is proved by Lyapunov method,and the effectiveness of the robust adaptive term is verified by the simulation.Experiments wearing the exoskeleton under different walking speeds and scenarios were carried out,and the results show that the absolute value of tracking errors of the hip and knee joints of the exoskeleton are consistently less than 1.5°and 2.5°,respectively.The proposed control algorithm effectively compensates for modeling errors and exhibits high robustness.

TCM-HIN2Vec:A strategy for uncovering biological basis of heart qi deficiency pattern based on network embedding and transcriptomic experiment

Objective:To elucidate the biological basis of the heart qi deficiency(HQD)pattern,an in-depth understanding of which is essential for improving clinical herbal therapy.Methods: We predicted and characterized HQD pattern genes using the new strategy,TCM-HIN2Vec,which involves heterogeneous network embedding and transcriptomic experiments.First,a heterogeneous network of traditional Chinese medicine(TCM)patterns was constructed using public databases.Next,we predicted HQD pattern genes using a heterogeneous network-embedding algorithm.We then analyzed the functional characteristics of HQD pattern genes using gene enrichment analysis and examined gene expression levels using RNA-seq.Finally,we identified TCM herbs that demonstrated enriched interactions with HQD pattern genes via herbal enrichment analysis.Results: Our TCM-HIN2Vec strategy revealed that candidate genes associated with HQD pattern were significantly enriched in energy metabolism,signal transduction pathways,and immune processes.Moreover,we found that these candidate genes were significantly differentially expressed in the transcriptional profile of mice model with heart failure with a qi deficiency pattern.Furthermore,herbal enrichment analysis identified TCM herbs that demonstrated enriched interactions with the top 10 candidate genes and could potentially serve as drug candidates for treating HQD.Conclusion: Our results suggested that TCM-HIN2Vec is capable of not only accurately identifying HQD pattern genes,but also deciphering the basis of HQD pattern.Furthermore our finding indicated that TCM-HIN2Vec may be further expanded to develop other patterns,leading to a new approach aimed at elucidating general TCM patterns and developing precision medicine.

Learning to Branch in Combinatorial Optimization With Graph Pointer Networks

Traditional expert-designed branching rules in branch-and-bound(B&B) are static, often failing to adapt to diverse and evolving problem instances. Crafting these rules is labor-intensive, and may not scale well with complex problems.Given the frequent need to solve varied combinatorial optimization problems, leveraging statistical learning to auto-tune B&B algorithms for specific problem classes becomes attractive. This paper proposes a graph pointer network model to learn the branch rules. Graph features, global features and historical features are designated to represent the solver state. The graph neural network processes graph features, while the pointer mechanism assimilates the global and historical features to finally determine the variable on which to branch. The model is trained to imitate the expert strong branching rule by a tailored top-k Kullback-Leibler divergence loss function. Experiments on a series of benchmark problems demonstrate that the proposed approach significantly outperforms the widely used expert-designed branching rules. It also outperforms state-of-the-art machine-learning-based branch-and-bound methods in terms of solving speed and search tree size on all the test instances. In addition, the model can generalize to unseen instances and scale to larger instances.

Graph Convolutional Networks Embedding Textual Structure Information for Relation Extraction

Deep neural network-based relational extraction research has made significant progress in recent years,andit provides data support for many natural language processing downstream tasks such as building knowledgegraph,sentiment analysis and question-answering systems.However,previous studies ignored much unusedstructural information in sentences that could enhance the performance of the relation extraction task.Moreover,most existing dependency-based models utilize self-attention to distinguish the importance of context,whichhardly deals withmultiple-structure information.To efficiently leverage multiple structure information,this paperproposes a dynamic structure attention mechanism model based on textual structure information,which deeplyintegrates word embedding,named entity recognition labels,part of speech,dependency tree and dependency typeinto a graph convolutional network.Specifically,our model extracts text features of different structures from theinput sentence.Textual Structure information Graph Convolutional Networks employs the dynamic structureattention mechanism to learn multi-structure attention,effectively distinguishing important contextual features invarious structural information.In addition,multi-structure weights are carefully designed as amergingmechanismin the different structure attention to dynamically adjust the final attention.This paper combines these featuresand trains a graph convolutional network for relation extraction.We experiment on supervised relation extractiondatasets including SemEval 2010 Task 8,TACRED,TACREV,and Re-TACED,the result significantly outperformsthe previous.

Neural network study of the nuclear ground-state spin distribution within a random interaction ensemble

The distribution of the nuclear ground-state spin in a two-body random ensemble(TBRE)was studied using a general classification neural network(NN)model with two-body interaction matrix elements as input features and the corresponding ground-state spins as labels or output predictions.The quantum many-body system problem exceeds the capability of our optimized NNs in terms of accurately predicting the ground-state spin of each sample within the TBRE.However,our NN model effectively captured the statistical properties of the ground-state spin because it learned the empirical regularity of the ground-state spin distribution in TBRE,as discovered by physicists.

Network Intrusion Detection Model Based on Ensemble of Denoising Adversarial Autoencoder

Network security problems bring many imperceptible threats to the integrity of data and the reliability of device services,so proposing a network intrusion detection model with high reliability is of great research significance for network security.Due to the strong generalization of invalid features during training process,it is more difficult for single autoencoder intrusion detection model to obtain effective results.A network intrusion detection model based on the Ensemble of Denoising Adversarial Autoencoder(EDAAE)was proposed,which had higher accuracy and reliability compared to the traditional anomaly detection model.Using the adversarial learning idea of Adversarial Autoencoder(AAE),the discriminator module was added to the original model,and the encoder part was used as the generator.The distribution of the hidden space of the data generated by the encoder matched with the distribution of the original data.The generalization of the model to the invalid features was also reduced to improve the detection accuracy.At the same time,the denoising autoencoder and integrated operation was introduced to prevent overfitting in the adversarial learning process.Experiments on the CICIDS2018 traffic dataset showed that the proposed intrusion detection model achieves an Accuracy of 95.23%,which out performs traditional self-encoders and other existing intrusion detection models methods in terms of overall performance.

Network pharmacology combined with molecular docking revealed the potential targets of Coridius chinensis in prostate cancer treatment

BACKGROUND Prostate cancer(PCa)has high morbidity and mortality rates in elderly men.With a history of thousands of years,traditional Chinese medicine derived from insects could be an important source for developing cancer-targeted drugs to prevent tumorigenesis,enhance therapeutic effects,and reduce the risk of recurrence and metastasis.Multiple studies have shown that Coridius chinensis(Cc)has anticancer effects.AIM To elucidate the mechanism of action of Cc against PCa via network pharma-cology and molecular docking.METHODS Potential targets for Cc and PCa were predicted using ChemDraw 19.0 software,the PharmMapper database and the GeneCards database.Then,the STRING database was used to construct the protein–protein interaction network.Gene Ontology(GO),Kyoto Encyclopedia of Genes and Genomes(KEGG)enrichment and molecular docking analyses were subsequently conducted to identify the key targets,active ingredients and pathways involved.RESULTS GO and KEGG analyses indicated that the PI3K-Akt signalling pathway was the critical pathway(P value<1.0×10-8).Multiple targeting ingredients that can affect multiple pathways in PCa have been identified in Cc.Seven active compounds(asponguanosines A,asponguanine B,asponguanine C,aspong-pyrazine A,N-acetyldopamine,aspongadenine B and aspongpyrazine B)were selected for molecular docking with 9 potential targets,and the results revealed that aspongpyrazine A and asponguanosine A are the main components by which Cc affects PCa(affinity<-5 kcal/mol,hydrogen bonding),but more studies are needed.CONCLUSION We used network pharmacology to predict the bioactive components and important targets of Cc for the treatment of PCa,supporting the development of Cc as a natural anticancer agent.

Predicting Users’ Latent Suicidal Risk in Social Media: An Ensemble Model Based on Social Network Relationships

Suicide has become a critical concern,necessitating the development of effective preventative strategies.Social media platforms offer a valuable resource for identifying signs of suicidal ideation.Despite progress in detecting suicidal ideation on social media,accurately identifying individuals who express suicidal thoughts less openly or infrequently poses a significant challenge.To tackle this,we have developed a dataset focused on Chinese suicide narratives from Weibo’s Tree Hole feature and introduced an ensemble model named Text Convolutional Neural Network based on Social Network relationships(TCNN-SN).This model enhances predictive performance by leveraging social network relationship features and applying correction factors within a weighted linear fusion framework.It is specifically designed to identify key individuals who can help uncover hidden suicidal users and clusters.Our model,assessed using the bespoke dataset and benchmarked against alternative classification approaches,demonstrates superior accuracy,F1-score and AUC metrics,achieving 88.57%,88.75%and 94.25%,respectively,outperforming traditional TextCNN models by 12.18%,10.84%and 10.85%.We assert that our methodology offers a significant advancement in the predictive identification of individuals at risk,thereby contributing to the prevention and reduction of suicide incidences.