Context: Complex networks represent systems with non-trivial connections and are widely used in fields such as social media, biology, and transportation. Temporal networks extend this by capturing the evolution of connections over time, providing insights into event sequences and information diffusion. Analyzing these networks requires specialized tools, and Python offers a variety of libraries tailored for this purpose.

Objectives: This study evaluates Python libraries designed for temporal network analysis based on multiple criteria. The aim is to assess the strengths and limitations of these tools, guide users in selecting appropriate libraries, and identify gaps for future development.

Methods: A comparative analysis was conducted on selected Python libraries using predefined evaluation criteria. The assessment considered factors such as available documentation, supported metrics, visualization capabilities, supported format, uniqueness, community support, and popularity. Data were gathered from official documentation, community forums, scientific papers, and usage statistics.

Results: Findings indicate that the TGX, Teneto, and PathpyG stand out, excelling in three of five criteria. Networkx-t shows balanced performance with no significant drawbacks, making it a reliable general-purpose choice. However, several tools have limitations in specific areas, such as a lack of comprehensive documentation or advanced visualization features.

Conclusion: This review provides an overview of existing Python tools for temporal network analysis, offering insights into their capabilities and shortcomings. The results assist researchers and practitioners in selecting suitable libraries while highlighting areas for improvement and potential future developments in the field.

Analyzing unlabeled trajectory data remains challenging due to high dimensionality, complex interactions between movement variables, and the lack of semantic annotations. In this paper, we present TaxVA, an interactive visual analytics system for exploring and interpreting spatio-temporal trajectories through a taxonomy-driven workflow. TaxVA organizes movement variables into semantically meaningful categories and integrates taxonomy-guided outlier analysis, zone-based behavioral partitioning, and feature-importance modeling within a coordinated visual interface. The system enables users to iteratively explore movement data by selecting analytical dimensions, inspecting behavioral zones, comparing representative trajectories, and linking statistical differences to spatial patterns. Through an interactive demonstration on real-world trajectory data, we show how TaxVA supports multi-perspective analysis, allowing the same trajectories to be interpreted in terms of both kinematic properties (e.g., speed variability) and geometric characteristics (e.g., turning behavior). The demo highlights how the integration of feature importance, temporal heatmaps, and spatial visualization facilitates intuitive and interpretable exploration of complex mobility data without requiring predefined labels.

E-commerce marketplace platforms rely increasingly on digital advertising to drive merchant sales, yet existing analytics tools often fail to provide contextual insights that connect advertising performance with actionable business strategies. This paper introduces ECommVis, a visual analytics platform designed to address the gap between complex advertising metrics and merchant decision-making in e-commerce marketplaces. ECommVis integrates five complementary visualization components that provide temporal, spatial, and comparative perspectives on advertising performance: Orders Over Time (OS) tracks historical organic versus ad-driven sales through adjustable time frames; Geospatial Interactions (GS) visualizes spatial performance patterns using hexagonal grid mapping; Merchant Metrics (MC) displays advertising KPIs via interactive stacked charts correlating budget with engagement; Interactions by Weekday and Shift (HM) identifies optimal advertising windows across daily shifts and weekday patterns; and Time Period Comparison (TP) benchmarks campaigns across time frames using correlation, trend, and search term analyses. We demonstrate the usefulness and effectiveness of ECommVis through two usage scenarios and a user study with 11 participants.

Forestry presents complex, irregular environments that are underrepresented in the visualization literature, despite the increasing availability of large-scale LiDAR and multispectral datasets. Extracting meaningful information from this data requires analyzing subtle structural and physiological variations among individual trees, a task that remains challenging due to the irregular, multi-layered nature of forests. In this paper, we present a visualization-driven approach for the interactive exploration and analysis of 3D forestry point clouds at the tree level, enriched with Normalized Difference Vegetation Index (NDVI) metrics. Our web-based system, entitled ForestVis, enables users to navigate dense forest environments, query individual tree attributes, and analyze vegetation metrics within their spatial context. The system design is informed by a semi-structured interview with forestry domain experts, from which we derive a set of analytical tasks emphasizing tree-centric abstractions, interactive exploration, and scalability to large datasets. Through illustrative use cases, we show how the system supports applications such as precision forestry, biodiversity monitoring, and detection of tree stress. More broadly, this work highlights the potential of interactive visualization providing novel tools and frameworks for ecological research and sustainable forest management. We evaluate ForestVis through two expert studies: a qualitative study of usability and analytical reasoning, and a quantitative study reporting objective task measures together with benchmarks of rendering, streaming, and interaction performance across two hardware configurations. Our evaluation suggested that our tool is effective in helping domain experts explore specific forest tasks, in large forest point cloud datasets, while maintaining computational performance. Our findings suggest that different point cloud rendering styles are suited to different tasks, whether analyzing individual trees or groups of trees, and interactive filtering helped experts relate vegetation indices to the point-cloud instance segmentation and to tree height.

Eye-tracking is widely used across disciplines, yet existing software solutions often require proprietary tools or programming expertise, creating barriers to exploratory gaze analysis. This paper presents Gazeful, a human-centered visual analytics system that integrates screen-based eye-tracking data acquisition with interactive visualization in a single open-source, graphical workflow, lowering the barrier to gaze analysis for users without programming expertise. The system enables recording, processing, and exploration of eye-tracked data through interactive heatmaps and gaze plots with adjustable analytical parameters and fixation-level inspection. Built in Python, Gazeful is designed to be extendable to new hardware and visual representations of gaze. Using well-established visualizations combined with interactive controls, Gazeful supports reproducible workflows and investigation of how visualization and interaction design influence eye-tracking analysis. A usability study with non-expert participants found the workflow to be effective and competitive with comparable graphical tools.

This paper presents the Tutorial Educational Program (TEP), an initiative designed to enhance academic engagement through integrated experiences in teaching, research, and community outreach. Inspired by the Brazilian Programa de Educação Tutorial (PET) and adapted to a Swedish higher education context, TEP offers undergraduate and master’s students structured opportunities to participate in departmental activities while developing critical, collaborative, and leadership skills. The program addresses challenges such as low retention in technology-related fields and employs a selective admissions process that emphasizes academic performance, motivation, and gender balance. Students work as part-time assistants under faculty supervision across three interconnected pillars. TEP follows a cyclical operational methodology of recruitment, participation, evaluation, and transition. This study evaluates TEP using multi-stakeholder survey and interview data, showing that participants perceive gains in pedagogical competence, academic identity, research understanding, and time-management skills. Teachers and administrators report increased reliability, continuity, and reduced coordination workload, while broader benefits for departmental teaching and outreach are identified. Early experiences indicate other benefits, including reduced administrative workload, continuous teaching and outreach support, and development of transferable skills. These findings highlight the potential of structured mentorship programs to support student development, research readiness, and institutional collaboration in higher education.

The need for large, high-quality annotated datasets continues to represent a primary limitation in training Object Detection (OD) models. To mitigate this challenge, we present VILOD, a Visual Interactive Labeling tool that integrates Active Learning (AL) with a suite of interactive visualizations to create an effective Human-in-the-Loop (HITL) workflow for OD annotation and training. VILOD is designed to make the AL process more transparent and steerable, empowering expert users to implement diverse, strategically guided labeling strategies that extend beyond algorithmic query strategies. Through comparative case studies, we evaluate three visually guided labeling strategies against a conventional automated AL baseline. The results show that a balanced, human-guided strategy—leveraging VILOD's visual cues to synthesize information about data structure and model uncertainty—not only outperforms the automated baseline but also achieves the highest overall model performance. These findings emphasize the potential of visually guided, interactive annotation to enhance both the efficiency and effectiveness of dataset creation for OD.

Analyzing Multivariate Networks (MVN) requires considering both graph structure and node attributes. However, most dimensionality reduction techniques operate on a single data modality, limiting their ability to comprehensively represent such networks. In this paper, we propose two workflows for MVN visualization that integrate structural and attribute information into a unified embedding process. The first workflow linearly combines structure- and attribute-based distances, while the second employs Graph Neural Network embeddings as an intermediate representation before projecting the data onto a 2D embedding. To evaluate neighborhood preservation across both data modalities, we introduce Harmonic Trustworthiness, a metric that balances structural and attribute-based preservation. Results on real-world datasets indicate improved representation and neighborhood preservation.

As manufacturing systems grow more complex, traditional 2D schematics and static models fall short in training and com-prehension. This work explores Augmented Reality (AR) as a solution, enabling interactive and spatially meaningful learningexperiences. Focusing on small and medium-sized enterprises (SMEs), the system integrates a Digital Twin (DT) with MicrosoftHoloLens and Unity to provide immersive visualization and interaction with mechanical machines. Users can manipulate com-ponents through natural hand gestures, enhancing spatial reasoning and hands-on training. This mixed-reality approach offersSMEs a scalable and practical tool for modernizing workforce education in manufacturing.

Manufacturing digitalization and automation research has expanded rapidly over the past decades. However, the current literature often presents fragmented views, lacking a comprehensive understanding of its evolution. Due to the vast number of publications, traditional literature reviews are impractical for broad fields like manufacturing digitalization and automation. This study leverages recent advancements in Artificial Intelligence (AI), Large Language Models (LLMs), and Natural Language Processing (NLP) to analyze 31,914 scientific papers from 1970 to 2023, providing a knowledge structure of the field. Moreover, we provide a roadmap of the field's evolution using Dynamic Topic Modeling (DTM). We note emerging trends in energy efficiency (since 2004), composite materials (2006), cybersecurity (2008), robotics (2014), and AI (2016), while simulation, scheduling, and process planning maintain steady and consistent research interests. Additionally, we observe a decay in certain research areas such as manufacturing automation protocol, a once-prominent area in the 1980s introduced by General Motors. We introduce a technology portfolio assessment framework categorizing technologies into Emerging, Established, Experimental, and Decaying quadrants. The proposed framework is based on the findings of this study, technology life cycle, creative destruction, and technological paradigms theories. The findings also offer insights into thematic shifts across the literature, consistent with creative destruction and technological paradigm theories. These findings hold implications for academia, industry, and policymakers, supporting more strategic innovation and resource allocation in the manufacturing sector.