Abstract
In many data stream applications, there is a normal concept, and the objective is to identify normal and abnormal concepts by training only with normal concept instances. This scenario is known in the literature as one-class learning (OCL) for data streams. In this OCL scenario for data streams, we highlight two main gaps: (i) lack of methods based on graph neural networks (GNNs) and (ii) lack of interpretable methods. We introduce OPENCAST (One-class graPh autoENCoder for dAta STream), a new method for data streams based on OCL and GNNs. Our method learns representations while encapsulating the instances of interest through a hypersphere. OPENCAST learns low-dimensional representations to generate interpretability in the representation learning process. OPENCAST achieved state-of-the-art results for data streams in the OCL scenario, outperforming seven other methods. Furthermore, OPENCAST learns low-dimensional representations, generating interpretability in the representation learning process and results.

Abstract
Influential nodes are the important nodes that most efficiently control the propagation process throughout the network. Among various structural-based methods, degree centrality, k-shell decomposition, or their combination identify influential nodes with relatively low computational complexity, making them suitable for large-scale network analysis. However, these methods do not necessarily explore nodes' underlying structure and neighboring information, which poses a significant challenge for researchers in developing timely and efficient heuristics considering appropriate network characteristics. In this study, we propose a new method (IC-SNI) to measure the influential capability of the nodes. IC-SNI minimizes the loopholes of the local and global centrality and calculates the topological positional structure by considering the local and global contribution of the neighbors. Exploring the path structural information, we introduce two new measurements (connectivity strength and effective distance) to capture the structural properties among the neighboring nodes. Finally, the influential capability of a node is calculated by aggregating the structural and neighboring information of up to two-hop neighboring nodes. Evaluated on nine benchmark datasets, IC-SNI demonstrates superior performance with the highest average ranking correlation of 0.813 with the SIR simulator and a 34.1% improvement comparing state-of-the-art methods in identifying influential spreaders. The results show that IC-SNI efficiently identifies the influential spreaders in diverse real networks by accurately integrating structural and neighboring information.

Abstract
Distributed Acoustic Sensing (DAS) leverages the sensitivity of optical fibers to detect environmental vibrations. This study demonstrates the capability of DAS to identify and characterize the acoustic signatures of passing vessels, highlighting its potential to enhance maritime surveillance and monitoring. © 2025 Elsevier B.V., All rights reserved.

Abstract

Abstract
This study focuses on improving the picking processes within a Picking-by-Line (PBL) warehouse through the development of a simulation model to assess different layouts and new operational rules. Utilizing a combination of Discrete Event Simulation (DES) and Agent-Based Modeling (ABS) in AnyLogic, the simulation model was validated against real-world Key Performance Indicators (KPIs) to ensure accuracy. The study identified three primary improvement opportunities. To address these opportunities, four scenarios were tested. The results showed varying impacts on productivity, with three of the four scenarios yielding improvements in picking productivity. Pilot testing confirmed the simulation model's predictions. The findings indicate that balancing travel distance reduction with congestion management is key to increasing picking productivity. This study reaffirms the value of simulation modeling in warehouse management, providing a robust framework for free-risk testing. © 2025 Elsevier B.V., All rights reserved.

Abstract
Latent Dirichlet Allocation (LDA) is a fundamental method for clustering short text streams. However, when applied to large datasets, it often faces significant challenges, and its performance is typically evaluated in domain-specific datasets such as news and tweets. This study aims to fill this gap by evaluating the effectiveness of short text clustering methods in a large and diverse e-commerce dataset. We specifically investigate how well these clustering algorithms adapt to the complex dynamics and larger scale of e-commerce text streams, which differ from their usual application domains. Our analysis focuses on the impact of high homogeneity scores on the reported Normalized Mutual Information (NMI) values. We particularly examine whether these scores are inflated due to the prevalence of single-element clusters. To address potential biases in clustering evaluation, we propose using the Akaike Information Criterion (AIC) as an alternative metric to reduce the formation of single-element clusters and provide a more balanced measure of clustering performance. We present new insights for applying short text clustering methodologies in real-world situations, especially in sectors like e-commerce, where text data volumes and dynamics present unique challenges.