Abstract:In the rapidly advancing information era, various human behaviors are being precisely recorded in the form of data, including identity information, criminal records, and communication data. Law enforcement agencies can effectively maintain social security and precisely combat criminal activities by analyzing the aforementioned data. In comparison to traditional data analysis methods, deep learning models, relying on the robust computational power in cloud centers, exhibit higher accuracy in extracting data features and inferring data. However, within the architecture of cloud centers, the transmission of data from end devices introduces significant latency, hindering real-time inference of data. Furthermore, low-latency edge computing architectures face limitations in direct deployment due to relatively weak computing and storage capacities of nodes. To address these challenges, a lightweight distributed knowledge graph completion architecture is proposed. Firstly, we introduce a lightweight distributed knowledge graph completion architecture that utilizes knowledge graph embedding for data analysis. Subsequently, to filter out substandard data, a personnel data quality assessment method named PDQA is proposed. Lastly, we present a model pruning algorithm that significantly reduces the model size while maximizing performance, enabling lightweight deployment. In experiments, we compare the effects of 11 advanced models on completing the knowledge graph of public security personnel information. The results indicate that the RotatE model outperforms other models significantly in knowledge graph completion, with the pruned model size reduced by 70\%, and hits@10 reaching 86.97\%.}

What problem does this paper attempt to address?

### Problems the Paper Attempts to Solve This paper aims to address the issues of anomaly detection and information completion in social network knowledge graphs, particularly in the application of public safety. Specifically, the paper focuses on the following aspects: 1. **Data Transmission Delay**: In traditional cloud computing architectures, transmitting data from terminal devices to the cloud center introduces significant delays, hindering the ability for real-time data inference. 2. **Limitations of Edge Computing**: Although low-latency edge computing architectures can reduce data transmission delays, the relatively weak computing and storage capabilities of nodes pose challenges for directly deploying deep learning models. 3. **Data Privacy and Security**: In the field of public safety, data privacy and security are crucial. Traditional cloud center architectures may face risks of data breaches due to centralized data storage. 4. **Construction and Optimization of Knowledge Graphs**: Effectively constructing and optimizing knowledge graphs to achieve efficient data analysis, anomaly detection, and information completion is an important research topic. To address these issues, the paper proposes a lightweight distributed knowledge graph completion architecture combined with edge computing technology, achieving the following goals: - **Data Structuring**: By constructing ontologies, extracting information, and integrating knowledge, data is transformed into structured knowledge, facilitating the quick and accurate retrieval of key information. - **Knowledge Graph Embedding (KGE)**: Utilizing KGE technology to infer missing knowledge in existing data, identify erroneous information, and perform anomaly detection and data completion. - **Model Pruning**: Significantly reducing model size through model pruning algorithms to improve performance, enabling lightweight deployment on edge devices. - **Real-time Processing and Data Privacy Protection**: Combining edge computing technology to ensure real-time processing of data on edge devices while protecting data privacy and security. Through these methods, the paper aims to provide law enforcement agencies with an efficient social management tool to help identify potential risks and improve public safety levels.

Edge-Enabled Anomaly Detection and Information Completion for Social Network Knowledge Graphs

Edge-enabled anomaly detection and information completion for social network knowledge graphs

Reconstructed Graph Neural Network With Knowledge Distillation for Lightweight Anomaly Detection

Accurate Anomaly Detection Leveraging Knowledge-enhanced GAT

Toward Anomaly Behavior Detection As an Edge Network Service Using a Dual-Task Interactive Guided Neural Network.

A Cybersecurity Knowledge Graph Completion Method for Scalable Scenarios

Optimized Edge-Cloud System for Activity Monitoring Using Knowledge Distillation

Extendable Multi-Device Collaborative Pipeline Parallel Inference in the Edge-Cloud Scenario

Continuous GNN-based Anomaly Detection on Edge using Efficient Adaptive Knowledge Graph Learning

KnowGraph: Knowledge-Enabled Anomaly Detection via Logical Reasoning on Graph Data

LLM-based Knowledge Pruning for Time Series Data Analytics on Edge-computing Devices

Exploring & Exploiting High-Order Graph Structure for Sparse Knowledge Graph Completion

FFKD-CGhostNet: A Novel Lightweight Network for Fault Diagnosis in Edge Computing Scenarios

Knowledge Based Attribute Completion for Heterogeneous Graph Node Classification

A Knowledge Graph Completion Algorithm Based on the Fusion of Neighborhood Features and vBiLSTM Encoding for Network Security

An Enhanced Group Abnormity Detection Model in Social Networks Through Multi-Scale Knowledge Graph-Based Deep Learning

Edge Computing and Few-shot Learning Featured Intelligent Framework in Digital Twin Empowered Mobile Networks

Hierarchical Perceptual Graph Attention Network for Knowledge Graph Completion

Adaptive Anomaly Detection for Internet of Things in Hierarchical Edge Computing: A Contextual-Bandit Approach

Cloud-Edge Collaborative Large Model Services: Challenges and Solutions

Dispelling the Fake: Social Bot Detection Based on Edge Confidence Evaluation