Light-Weight Graph Matching Query Over Encrypted Graphs

Dynamic Vehicle Route Planning (DVRP) severely challenges logistics system optimization as it must respond to dynamic events like unexpected orders and traffic changes in real time. Traditional methods suffer from high computational overhead and poor adaptability to dynamic environments, while existing deep learning methods struggle to balance solution quality and computational efficiency. To address this, this paper proposes the lightweight graph capsule network (LGCN) algorithm. It encodes dynamic path states into spatio-temporal graph structures, uses a hierarchical capsule mechanism instead of traditional graph convolution, and applies lightweight processing to the hierarchical capsule network. This effectively resolves the conflict between fast response to dynamic changes and high-quality computational results. The method's effectiveness is evaluated by comparison with transformer-based algorithms, OR-Tool algorithms, and Clarke Wright's algorithm. Experiments show that LGCN not only significantly improves real-time dynamic response efficiency but also performs excellently in multiple key metrics such as average routing cost and average vehicles.

The estimation of the camera poses associated with a set of images commonly relies on feature matches between the images. In contrast, we are the first to address this challenge by using objectness regions to guide the pose estimation problem rather than explicit semantic object detections. We propose Pose Refiner Network (PoserNet) a light-weight Graph Neural Network to refine the approximate pair-wise relative camera poses. PoserNet exploits associations between the objectness regions - concisely expressed as bounding boxes - across multiple views to globally refine sparsely connected view graphs. We evaluate on the 7-Scenes dataset across varied sizes of graphs and show how this process can be beneficial to optimisation-based Motion Averaging algorithms improving the median error on the rotation by 62\(^{\circ }\) with respect to the initial estimates obtained based on bounding boxes. Code and data are available at github.com/IIT-PAVIS/PoserNet.

Attention Deficit Hyperactivity Disorder (ADHD) is a behavior-based disorder that mainly occurs in young children. Resting-state fMRI data have been very popular for diagnosing brain disorders like Autism, ADHD, and schizophrenia, by network-based functional connectivity, since these disorders are associated with both individual brain regions and their connections. Finding patterns among regions of controls’ brain and ADHD patients’ discriminating brains, is a non-trivial task. For classification of ADHD, we propose an end-to-end lightweight CNN architecture with hierarchical representation learning i.e., HLGSNet. We extract 116 anatomical regions from each subject in both normal and patient conditions, and graphs are built with the help of temporal correlation between different regions, where each region is considered as a node. Following this, a Siamese graph convolution neural network with triplet loss has been trained for finding embeddings so that samples for the same class should have similar embeddings. Finally, along with a fully connected layer, the trained model has been fine-tuned for the classification task. Experiments have been carried out on publicly available ADHD-200 dataset with promising performance.

Graph convolutional network (GCN) has been widely used by skeleton based action recognition algorithms and achieves remarkable performance. However, recent GCN based State-Of-The-Art (SOTA) models for skeleton based action recognition tend to become increasingly sophisticated and over-parameterized. The low efficiency in model training and inference poses a challenge for their practical implementation in real-world scenarios. To address this issue, we construct a GCN based lightweight model for skeleton based action recognition, termed LightGCN. In this work we introduce an efficient convolutional neural network (CNN) structure to our temporal convolutional (TC) layer to extract temporal dynamics, effectively reducing model complexity. Furthermore, we propose a novel attention module that first extends the multi-spectral channel attention mechanism to the field of skeleton based action recognition, which preserves not only the lowest frequency information, but also useful information encoded by other frequency components, reducing the information loss during the channel compression. In order to further reduce the model complexity, we design a new compound scaling strategy to expand the model’s width and depth to different extent. This strategy enables the model to achieve an excellent balance between complexity and accuracy. On the two large-scale datasets, i.e., NTU RGB+D 60 and 120, our proposed LightGCN achieves 92.5% accuracy on the cross-subject benchmark of NTU 60 dataset, outperforming previous SOTA lightweight models and most heavyweight models, while needing 24.54% fewer parameters and 24.76% fewer flops than EfficientGCN-B4, which is the SOTA lightweight model.

Botnets represent a significant threat to the security of the current internet network environment. They can be employed to carry out various malicious activities, not only impairing the performance and security of individual devices but also causing extensive damage to the entire network. Therefore, efficient detection of botnets is essential for network security. In recent years, the mainstream approach to botnet detection has involved the use of graph neural network (GNN) models, aiming to fully utilise the graph-structured properties of network data. However, current GNN methods frequently struggle with complicated structures and lack interpretability. To address these issues, we have designed a new model based on graph isomorphism networks (GINs). This model significantly simplifies the complexity without compromising detection accuracy and introduces explainable techniques for analysing model weights and conducting subgraph mining. By leveraging the concept of graph isomorphism, our method can more precisely characterise and identify botnet features. Experimental results demonstrate that our model excels in both accuracy and interpretability, making it highly valuable for real-world network environment applications.

LightGCAN: A lightweight graph convolutional attention network for user preference modeling and personalized recommendation

We propose LightNE, a cost-effective, scalable, and high quality network embedding system that scales to graphs with hundreds of billions of edges on a single machine. In contrast to the mainstream belief that distributed architecture and GPUs are needed for large-scale network embedding with good quality, we prove that we can achieve higher quality, better scalability, lower cost and faster runtime with shared-memory, CPU-only architecture. LightNE combines two theoretically grounded embedding methods NetSMF and ProNE. We introduce the following techniques to network embedding for the first time: (1) a newly proposed downsampling method to reduce the sample complexity of NetSMF while preserving its theoretical advantages; (2) a high-performance parallel graph processing stack GBBS to achieve high memory efficiency and scalability; (3) sparse parallel hash table to aggregate and maintain the matrix sparsifier in memory; and (4) Intel MKL for efficient randomized SVD and spectral propagation.

Shared-account Sequential Recommendation (SSR) aims to provide personalized recommendations for accounts shared by multiple users with varying sequential preferences. Previous studies on SSR struggle to capture the fine-grained associations between interactions and different latent users within the shared account's hybrid sequences. Moreover, most existing SSR methods (e.g., RNN-based or GCN-based methods) have quadratic computational complexities, hindering the deployment of SSRs on resource-constrained devices. To this end, we propose a Lightweight Graph Capsule Convolutional Network with subspace alignment for shared-account sequential recommendation, named LightGC2N. Specifically, we devise a lightweight graph capsule convolutional network. It facilitates the fine-grained matching between interactions and latent users by attentively propagating messages on the capsule graphs. Besides, we present an efficient subspace alignment method. This method refines the sequence representations and then aligns them with the finely clustered preferences of latent users. The experimental results on four real-world datasets indicate that LightGC2N outperforms nine state-of-the-art methods in accuracy and efficiency.

Effective channel estimation (CE) is critical for optimizing the performance of 5G New Radio (NR) systems, particularly in dynamic environments where traditional methods struggle with complexity and adaptability. This paper introduces GraphNet, a novel, lightweight Graph Neural Network (GNN)-based estimator designed to enhance CE in 5G NR. Our proposed method utilizes a GNN architecture that minimizes computational overhead while capturing essential features necessary for accurate CE. We evaluate GraphNet across various channel conditions, from slow-varying to highly dynamic environments, and compare its performance to ChannelNet, a well-known deep learning-based CE method. GraphNet not only matches ChannelNet’s performance in stable conditions but significantly outperforms it in high-variation scenarios, particularly in terms of Block Error Rate. It also includes built-in noise estimation that enhances robustness in challenging channel conditions. Furthermore, its significantly lighter computational footprint makes GraphNet highly suitable for real-time deployment, especially on edge devices with limited computational resources. By underscoring the potential of GNNs to transform CE processes, GraphNet offers a scalable and robust solution that aligns with the evolving demands of 5G technologies, highlighting its efficiency and performance as a next-generation solution for wireless communication systems.

A lightweight fork of 'gMCP' with functions for graphical described multiple test procedures introduced in Bretz et al. (2009) &lt;<a href="https://doi.org/10.1002%2Fsim.3495" target="_top">doi:10.1002/sim.3495</a>&gt; and Bretz et al. (2011) &lt;<a href="https://doi.org/10.1002%2Fbimj.201000239" target="_top">doi:10.1002/bimj.201000239</a>&gt;. Implements a flexible function using 'ggplot2' to create multiplicity graph visualizations. Contains instructions of multiplicity graph and graphical testing for group sequential design, described in Maurer and Bretz (2013) &lt;<a href="https://doi.org/10.1080%2F19466315.2013.807748" target="_top">doi:10.1080/19466315.2013.807748</a>&gt;, with necessary unit testing using 'testthat'.

Campus violence action recognition based on lightweight graph convolution network

This paper proposes the two different approaches to speed-up program build: making link-time optimization work in parallel and lightweight optimization approach. The former technique is achieved by scaling LTO system. The latter makes link to work faster because of using summaries to manage some of interprocedural optimization passes instead of full IR code in memory. The problem of horizontal LTO system scaling leads to the problem of partition of the large task to several subtasks that can be performed concurrently. The problem is complicated by the compiler pipeline model: interprocedural optimization passes works consequentially and depends on previous performed ones. That means we can divide just data on which passes works, not passes themselves. We need to divide IR code to sub-independent parts and run LTO on them in parallel. We use program call graph analysis to divide a program to parts. Therefore, our goal is to divide call graph that is one of NP-compete problems. Nevertheless, the choice of the dividing algorithm strongly depends on properties of divided graph. The main goal of our investigation is to find lightweight graph partition algorithm that works efficiently on call graphs of real programs and that does not spoil LTO performance achievements after optimizing code pieces separately. This paper proposes new graph partition algorithm for program call graphs, results of comparing this one with two other methods on SPEC CPU2000 benchmark and implementation of the algorithm in scalable LLVM-based LTO system. The implementation of this approach in LTO system shows 31% link speedup and 3% of performance degradation for 4 threads. The lightweight optimization shows 0,5% speedup for single run in lazy code loading mode.

Facing the real dilemma of scarce social media attention and the homogenization of city brand symbols, this paper focuses on the Guangdong-Hong Kong-Macao Greater Bay Area and explores how to simultaneously improve brand visual consistency and communication effectiveness through a data-driven approach. To this end, we propose a technical framework that couples controllable diffusion generation with lightweight graph learning. On the generation side, Stable Diffusion is used as the backbone, combined with ControlNet and LoRA, to achieve controllable generation of landmark structures and CI color gamuts. On the communication side, LightGCN is used to model the “user-content” bipartite graph, and cross-modal attitude metrics are combined to construct a closed loop for candidate screening and delivery optimization. Experimental evaluation shows that the proposed method achieves stable improvements in CLIP alignment, brand consistency, and cultural alignment metrics, while significantly outperforming rule-based delivery and traditional graph recommendation in communication metrics such as interaction rate, forwarding rate, and second-order diffusion depth. Parameter sensitivity analysis further provides reasonable ranges for embedding dimension and time decay. This paper's innovation lies in proposing an integrated, verifiable paradigm for “generation quality—brand specifications—audience diffusion,” building an interpretable BCI/CCI indicator system and a coordinated selection mechanism, and providing a reusable engineering path and evaluation benchmark for the large-scale production and refined dissemination of city brand AIGC.

Purpose Offline commercial complexes feature diverse services, sparse behavioural data and rapidly changing user interests, which limit conventional recommendation methods. This study develops a personalised recommendation framework designed for high-uncertainty offline environments and evaluates its performance across static and dynamic systems and across cold-start and stable user groups. Design/methodology/approach A lightweight system is built by encoding questionnaire data into a preference knowledge graph and applying a hybrid user similarity model for preference inference. A dynamic module updates user–category relations through explicit feedback and controlled semantic diffusion, enabling continuous refinement. Static and dynamic configurations are tested separately on cold-start and stable users and ablation experiments quantify the contribution of relation updating and diffusion. Findings The static system provides a strong baseline under sparse data. The dynamic system yields further improvements: for cold-start users, P@1 and HIT@3 rise by 5.6 and 6.6 percentage points; for stable users, P@1 increases by 10.0 percent with clear gains in mean reciprocal rank and discounted cumulative gain. Ablation results show that feedback-driven edge updating generates most improvements, while intra-cluster diffusion supplies additional benefits, particularly for cold-start users. Originality/value The study offers a structurally adaptable and computationally efficient framework for offline commercial environments without behavioural logs. Through a combined static-to-dynamic and cold-start-to-stable evaluation, it demonstrates how lightweight graph modelling and local feedback-based updates can deliver accurate, interpretable and deployable personalised recommendations for real facility and commercial management settings.

Optimizing fault tree traversal order for efficient minimal cut set generation: A lightweight graph neural network approach