Vol. 44 · No. 3

Aspect-Based Sentiment Analysis (ABSA) aims to predict the sentiment polarity of the given aspect word within the sentence. Recent studies frequently treat syntactic and semantic features as independent representations, thereby overlooking their intrinsic correlation. Concurrently, most of the existing methods largely neglect the significance of dependency types, which eventually impacts the accuracy of sentiment analysis. Research based on cognitive theory indicates a mutual influence between syntax and semantics. Based on this, we propose an ABSA model based on enhanced multi-feature graph convolutional network (Triple-GCN). Firstly, a shared enhanced graph convolutional module is proposed to integrate syntactic and semantic information. Following this, a thorough fusion of this syntactic and semantic information is carried out. Besides, relation and adjacency matrices are utilized for the innovative reconstruction of hidden state vectors. Syntactic graph convolution module dynamically fuses hidden state vectors and dependency features. Additionally, a position weight encoding function is designed to comprehend sentiment dependencies by drawing attention to aspect-near words. On the semantic side, dynamic semantic graphs are constructed, enabling the capture of semantic features. The model has been evaluated on three public datasets: Twitter, Laptop14, and Restaurant14. Compared to existing baseline models, the effectiveness of this model has noticeably improved.

In this work, we start by demonstrating experimentally that rhetorical question detection is still a challenging task, even for the state-of-the-art Large Language Models (LLMs). We propose an approach that boosts the performances of such LLMs by training a soft prompt in a way that enables building a joint embedding space from multiple loosely related corpora. The advantages of using a soft-prompt compared to fine-tuning is to limit the training costs and combat overfitting and forgetting. Soft prompting is often viewed as a way to guide the model towards a specific known task, or to introduce new knowledge into the model through in-context learning. We show that soft prompting may also be used to modify the geometry of the embedding space, so that the distance between embeddings becomes semantically relevant for a target task, similarly to what is commonly achieved with contrastive fine-tuning. We exploit this property to combat data scarcity for the task of rhetorical question detection by merging several datasets into a joint semantic embedding space. On the standard Switchboard dataset we demonstrate that the resulting BERT-based model nearly divides by 2 the number of errors as compared to Flan-T5-XXL with only 5 few-shot labeled samples, thanks to this joint embedding space. We have chosen in our experiments a BERT model because it has already been shown with S-BERT that contrastive fine-tuning of BERT leads to semantically meaningful representations. Therefore, we also show that this property of BERT nicely transfers to the soft-prompting paradigm. Finally, we qualitatively analyze the resulting embedding space and propose a few heuristic criteria to select appropriate related tasks for inclusion into the pool of training datasets.

In view of the dependence of existing 3D action prediction research on labels, we propose a graph convolutional recurrent 3D action prediction method based on state discrimination and spatio-temporal self-supervised contrast learning. In the state discrimination task, cross-sample sampling and relative action completeness perception are used to train the model for generalized state information learning across instances and classes. In the spatio-temporal contrast task, spatio-temporal consistency information is introduced into the feature representation to enrich action semantics in features. Additionally, in order to fully extract spatio-temporal information in 3D action sequences, a spatio-temporal feature extraction network (STFEN) based on graph convolution recurrent network is designed. The experimental results on public datasets demonstrate the efficiency of the proposed methods.

As the demand for data privacy protection continues to grow and the concept of collaborative modeling gains traction, federated learning has emerged as a pivotal distributed learning paradigm in the Internet of Things (IoT) domain. However, the client data held by different institutions often varies significantly in sources and characteristics, which can hinder the efficiency of federated learning model training and increase the risk of personal privacy breaches. To address the challenges of model accuracy degradation and privacy exposure when federated learning is applied to multi-source heterogeneous data, we propose a personalized federated learning strategy that integrates hypernetworks with attention mechanisms. This strategy involves transforming labeled data at the source to protect personal privacy while employing hypernetworks and Transformer-based mechanisms to focus on the personalized information of clients from various institutions. Our proposed approach supports handling heterogeneous data, thereby better meeting the personalized needs of different institutions. Experimental results demonstrate that this framework not only effectively safeguards data privacy but also significantly enhances the performance and generalization capability of federated learning on heterogeneous data. This research offers a novel perspective for developing more adaptable personalized federated learning models, facilitating cross-institutional collaborative research, and providing an innovative model training solution for various IoT devices, balancing the dual requirements of data privacy protection and multi-institutional data sharing.

Due to the substantial volume of data associated with the IoT, processing and storing such large amounts of data is not easily feasible. Nevertheless, many of its applications face challenges in cloud computing, such as latency, location awareness, and real-time mobility support. Edge computing helps provide solutions to these challenges. In this article, the MINLP path optimization problem is initially addressed using SDN, SA+GA, OLB-LBMM, and Round-Robin methods. Subsequently, based on the obtained results, the SDN method, which has achieved the best outcomes among the approaches, is selected. This article involves a simulation of IoT for optimal allocation of shared resources in edge computing. The network architecture comprises five distinct layers, including cloud services, the SDN controller, edge computing nodes, edge computation and users. The algorithm employed in this problem is the federated learning and stochastic gradient descent algorithm. It selects the optimal edge node for user service provision through two learning and training phases, aiming to allocate shared resources to three parameters: cloud service providers' revenue, average latency, and user satisfaction. This algorithm is compared with several other methods. The selected model and algorithm, in comparison with other algorithms used in solving similar models, lead to a centralized management system, the implementation of effective network management, and the utilization of various communication media. This approach ensures timely access to services, contributing to increased profits for providers and user satisfaction.

In the current research on heterogeneous academic network community detection, there is a widespread challenge of high demand for node representation of node attributes in learning graphs. Particularly, existing methods often perform poorly when dealing with nodes missing attributes. Furthermore, most methods rely on meta-paths, but the optimal length of meta-paths is difficult to determine and the quality of predefined meta-paths directly affects the results. To address this issue, this paper proposes an Adaptive Sampling-based Heterogeneous Graph Enhancement Model (ASGNN). The model aims to solve the problem of inaccurate node representations leading to imprecise community partitions in academic networks. ASGNN first effectively captures the network's topological structure through random walk techniques, and then utilizes an adaptive sampling algorithm to select the most influential adjacent node set, rather than relying on traditional meta-path techniques. The model further employs an attention mechanism to aggregate information from nodes of different types, thereby enhancing attribute completion and topological structure in heterogeneous academic networks. This approach not only fills in missing information but also significantly enhances the semantic and structural integrity of the network. Experimental results demonstrate that the proposed model exhibits outstanding performance on two real datasets compared to baseline models.

Using convolutional neural networks and genetic programming, this study presents a new composite technique for modeling bicycle traffic in the town of Novo mesto, Slovenia. Every town needs public passenger transportation because the current transportation system has well-known issues like congestion, environmental effect, a lack of parking spaces, increased safety hazards, and excessive energy consumption. Urban transport is crucial for the functionality of any city. High-quality and usable urban transport not only affects the functionality of the city as an economic and social center, but it also reduces the number of passenger cars on the streets. The Novo mesto region, which has a population of around 30 000 people, is a major industrial center that is strongly reliant on metropolitan transportation. Unfortunately, the urban traffic of Novo mesto still has a relatively weak influence on the transport connectivity of the wider area. The study's goal is to examine and simulate bicycle rentals. For 35 weeks, convolutional neural networks and genetic programming were utilized to anticipate bicycle traffic. Three types of models were applied to study the impact of weather conditions on bicycle traffic: linear regression, genetic programming, and feed-forward neural networks. The proposed approach will be useful for cities with similar needs around the world.

Federated learning (FL) allows distributed devices to jointly train a global model while safeguarding the privacy of their local data. However, selecting and securing clients, especially in environments with potentially malicious participants, remains a critical challenge. This study proposes an innovative participant selection method to enhance both security and efficiency in centralized and decentralized FL frameworks. In the centralized framework, this method effectively excludes clients with weak privacy protections and optimization capabilities, thus increasing overall system security. For decentralized FL, a blockchain-supported approach is introduced, which further strengthens the robustness of the system. Using a dynamic role assignment algorithm, roles such as worker, validator, and miner are allocated based on security and performance metrics for each training round. The findings show that this method performs on a par with the scenarios free of malicious clients, demonstrating the value of blockchain technology in improving FL protocols. By addressing security vulnerabilities and improving training efficiency, this research contributes to the development of more secure and efficient FL systems, underscoring the importance of advanced participant selection and role assignment strategies.

Knowledge graph representation learning aims to embed entities and relationships into low-dimensional space through knowledge graph embedding methods. Because knowledge graphs are incomplete, it is often necessary to complete the knowledge graph through representation learning methods. With the development of pre-trained language models, more and more research applies them to the field of knowledge graph representation learning, using the powerful semantic representation capabilities of pre-trained language models to improve the performance of knowledge graph embedding. Most of the existing methods make use of the semantic information of the triple text but do not fully consider the structural information of the triple and the graph structure information of the knowledge graph. The triple structure reflects the semantic information and relationship pattern of the triple, and the graph structure reflects the surrounding entity's semantic features. To address the above issues, this paper proposes a knowledge graph representation learning method named PREGSE, which is based on pre-trained language models and integrates graph structure information. Firstly, pre-trained language models are employed to encode triplets through text encoding, obtaining vectors for the triplets. Secondly, a graph attention network is utilized to learn various local graph structure information. Lastly, a multi-task learning strategy is applied to simultaneously learn triplet structure information and semantic information. We trained our model on the FB15k-237 and WN18RR datasets, and the results show that on the FB15k-237 dataset, our model improved the MRR metric by 27

Health care sector faces major challenges in light of the appearance of new diseases. The emergency department (ED) is an important area in the hospital, where it plays a major role in the presence of these challenges. The `ED' is a complex system due to the random flow of patients and the complex nature of its resources. In this paper, the authors model and simulate the `ED' by means of coloured Petri nets, and to determine the appropriate amount of resources, the NSGA-II algorithm is developed. After determining the appropriate amount of resources through the NSGA-II algorithm, the simulation model of the current system is modified with the amount of new resources obtained through the NSGA-II algorithm. The results are compared between the current system and the obtained system. This study was conducted in Hassani Abdelkader Hospital, located in the city of Sidi Bel Abbes, in western Algeria.