Search Results (183)

Handwritten Keyword Spotting (KWS) remains a challenging task, particularly in segmentation-free scenarios where word images must be retrieved and ranked based on their similarity to a query without relying on prior page-level segmentation. Traditional KWS methods primarily focus on visual similarity, often overlooking the underlying semantic relationships between words. In this work, we propose a novel NLP-driven re-ranking approach that refines the initial ranked lists produced by state-of-the-art KWS models. By leveraging semantic embeddings from pre-trained BERT-like Large Language Models (LLMs, e.g., RoBERTa, MPNet, and MiniLM), we introduce a relevance feedback mechanism that improves both verbatim and semantic keyword spotting. Our framework operates in two stages: (1) projecting retrieved word image transcriptions into a semantic space via LLMs and (2) re-ranking the retrieval list using a weighted combination of semantic and exact relevance scores based on pairwise similarities with the query. We evaluate our approach on the widely used George Washington (GW) and IAM collections using two cutting-edge segmentation-free KWS models, which are further integrated into our proposed pipeline. Our results show consistent gains in Mean Average Precision (mAP), with improvements of up to $2.3\%$ (from $94.3\%$ to $96.6\%$) on GW and $3\%$ (from $79.15\%$ to $82.12\%$) on IAM. Even when mAP gains are smaller, qualitative improvements emerge: semantically relevant but inexact matches are retrieved more frequently without compromising exact match recall. We further examine the effect of fine-tuning transformer-based OCR (TrOCR) models on historical GW data to align textual and visual features more effectively. Overall, our findings suggest that semantic feedback can enhance retrieval effectiveness in KWS pipelines, paving the way for lightweight hybrid vision-language approaches in handwritten document analysis. Full article

This paper studies Google’s PageRank algorithm. By an innovative application of the method of gluing Markov chains, we study the properties of Markov chains and extend their applicability by accounting for the damping factor and the personalization vector. Many properties of Markov chains related to spectrums and eigenvectors of the transition matrix, including the stationary distribution, periodicity, and persistent and transient states, will be investigated as well as part of the gluing process. Using the gluing formula, it is possible to decompose a large network into some sub-networks, compute their PageRank separably and glue them together. The computational workload can be reduced. Full article

As the global energy structure transitions towards cleaner sources, large-scale integration of wind power has become a trend for modern power systems. However, the impact of low-inertia power electronic converters and the fault propagation effects at critical nodes pose significant challenges to power system stability. To this end, a Kuramoto-like model analysis method, considering node importance, is proposed in this paper. First, virtual node technology is utilized to optimize the power grid topology model. Then an improved PageRank algorithm embedded by a critical node identification method is proposed, which simultaneously considers transmission efficiency, coupling transmission probability, and voltage influence among nodes. On this basis, the traditional uniform coupling assumption is eliminated, thereby reallocating the coupling strength between critical nodes. In addition, the Kron method is applied to simplify the power grid model, constructing a hybrid Kuramoto-like model that integrates second-order synchronous machine oscillators and first-order wind power oscillators. Based on this model, the transient stability of the wind power integrated power system is analyzed. Finally, through estimating the attraction region range of the stable equilibrium point, a transient stability criterion is proposed for fault limit removal time assessment. The simulation results of the improved IEEE 39-bus system show that coupling strength optimization based on node importance reduces the system’s average critical coupling strength by 17%, significantly improving synchronization robustness. Time-domain simulations validate the accuracy of the method, with the relative error of fault removal time estimation controlled within 10%. This research provides a new analytical tool for transient stability analysis of wind power integration. Full article

This study examines the applicability of the well-established PageRank algorithm for ranking teams and predicting outcomes in incomplete, single-elimination high school baseball tournaments. Using match data from Taiwan’s CTBC Black Panther Cup National High School Baseball Tournament spanning from 2013 to 2023, this research investigates whether PageRank can produce valid, stable, and predictive rankings under structural constraints and limited data environments. Three empirical evaluations were conducted. First, a comparative analysis between PageRank rankings and official results demonstrated high ordinal consistency, with Kendall’s tau values exceeding 0.70 in most seasons. Second, PageRank rankings were assessed for temporal robustness, demonstrating stable performance across seasons and under varying data inputs. Third, a series of n-step-ahead simulations were implemented to test the predictive validity of PageRank. The results indicate that incorporating historical data substantially improves forecasting accuracy, achieving up to 92.9% when data from up to four previous seasons are included. Overall, the findings support PageRank as a consistent and interpretable ranking method that is well-suited for grassroots sports. Its ability to incorporate indirect competition and opponent strength makes it effective in settings with sparse or unbalanced schedules. This study provides methodological insights and practical implications for ranking and evaluation in school-level sports. Full article

This research investigates the application of search engine optimization (SEO) in developing the digital image of rural tourism businesses in the Terras de Trás-os-Montes region of Portugal. With digital marketing becoming increasingly important for businesses to stay competitive, SEO has become a vital tool for developing online recognition, qualified traffic acquisition, and enhancement of conversion rates. The research performs an SEO analysis of 21 rural tourism websites by applying the Ubersuggest tool, analyzing such key indicators as on-page SEO scores, organic traffic, keyword ranking, backlinks, and technical performance. The results identify wide SEO performance discrepancies, with some sites registering excellent practices and others with critical errors that impair the sites’ online recognizability. In particular, low word count, absent meta description, and loading speed issues are very much present. The research emphasizes the need for effective SEO methods, such as on-page maintenance, content creation, and link building, to advance search engine ranking and end-user experience. Moreover, the study emphasizes the necessity for rural tourism businesses to evolve and adapt to current SEO trends, i.e., voice search optimization and local SEO, in the changing digital business environment. The results provide recommendations for rural tourism businesses to develop their digital marketing activities and make progress online. Full article

This study investigates the relationship between intercity transportation accessibility and network centrality across South Korea by integrating Global Positioning System (GPS)-based mobility data with graph-theoretic centrality measures, including degree, PageRank, local clustering coefficient, harmonic, Katz, and information centrality. Employing both statistical modeling and machine learning techniques, this analysis uncovers key structural patterns and interaction effects within the national mobility network. The findings yield several important insights. First, the Seoul Metropolitan Area emerges as the dominant mobility hub, with Busan, Daegu, and Daejeon functioning as secondary centers, reflecting a polycentric urban configuration. Second, intermediary transfer hubs—despite having lower direct connectivity—substantially enhance overall network efficiency and interregional mobility. Third, transportation accessibility, particularly in relation to regional transit and highway infrastructure, exhibits a significant association with centrality measures and strong feature importance, identifying these modes as primary determinants of spatial connectivity. Fourth, the impact of accessibility on centrality is characterized by nonlinear relationships and threshold effects. By elucidating the complex interplay between mobility infrastructure and spatial network dynamics, this study contributes to a more comprehensive understanding of regional connectivity and network centrality and offers policy-relevant insights for future transportation planning. Full article

Community detection is increasingly important in social networks with the rapid growth of big data, which provides a deep understanding of the mesoscopic structure of social networks. In this article, we propose a label improvement algorithm, DegreeRank Label Propagation (DRLP), which is based on the degree centrality of nodes and adopts a PageRank optimization strategy. We present a damping factor reflecting the affinity between nodes, which can be adjusted to affect the change of affinity between nodes caused by unexpected events, aiming to simulate interpersonal communication in real networks. Next, a novel importance index is designed for nodes to solve the random problem of existing similar algorithms by globalizing the local characteristics of nodes. We also develop an update algorithm with low time complexity during the label selection process to ensure the sum of influence propagation is maximized within each community. Experimental results verify that the algorithm achieves stable and excellent community partitioning results on real network datasets and artificial synthetic networks. Especially in large and medium-sized networks, our method demonstrates higher accuracy and better performance in terms of normalized mutual information (NMI) and modularity than other methods. Full article

Railway stations serve as critical nodes within transportation networks, and the efficient management of in-station track sections is vital for smooth operations. This study proposes an integrated method for identifying critical track sections, which refers to track sections with the highest static occupancy rates (HiSORTS), in railway station yards using a multiplex network framework. By modeling the station as a Railway Station Multiplex Network (RSMN) that incorporates train routes (TRs), extended routes (ERs), and shunting routes (SRs), the proposed approach overcomes the limitations of single-layer, single-metric analyses and effectively captures complex operational characteristics. Classical network metrics, including Degree Centrality (DC), Betweenness Centrality (BC), Closeness Centrality (CC), Katz Centrality (KC), and PageRank (PR), along with a custom Fusion Centrality (FC), are used to quantify track section importance. Principal Component Analysis (PCA) and the Technique for Order Preference by Similarity to Ideal Solution (TOPSIS) are applied to generate rankings, which are further analyzed using SHapley Additive exPlanations (SHAP)-based matrics contributions analysis. The results indicate that TR metrics contribute the most (50.3%), followed by ER (25.5%) and SR (24.2%), with KC and FC being the most influential metrics. The findings provide a robust decision-support framework for railway operations, facilitating targeted maintenance, congestion mitigation, and efficiency optimization. Full article

The objective of this paper is to demonstrate the structural characteristics of the European gas supply system and changes in its network structure and the interaction and clustering among its nodes defined as countries, following the outbreak of the Russia–Ukraine conflict. The methodology relies on social network analysis, which employs mathematics of the graph theory to examine the state and dynamics of the given network structure. The impacts identified involve the reduced reliance of the system on Russian gas, a considerable reduction in the strength centrality of Russia and Germany, and a higher dispersion of gas flows, largely due to the increased import of LNG flows. After the conflict outbreak, countries such as Italy, Austria, the Slovak Republic, and Hungary retained their high influential position, in terms of the PageRank centrality, while the Balkan countries, together with the Middle East gas suppliers (Turkey and Iran), formed a common group with Russia. The estimated changes stress the challenges posed to the EU countries to enhance connectivity infrastructure investments and explore alternative ways of gas supply to support the objectives of strategic autonomy, while promoting resilience and the path toward a carbon-free transition. Full article

Real-time anomaly detection in large, dynamic graph networks is crucial for real-world applications such as network intrusion prevention, fraud transaction identification, fake news detection in social networks, and uncovering abnormal communication patterns. However, existing graph-based methods often focus on static graph structures, which struggle to adapt to the evolving nature of these graphs. In this paper, we propose Adaptive-DecayRank, a real-time and adaptive anomaly detection model for dynamic graph streams. Our method extends the dynamic PageRank algorithm by incorporating an adaptive Bayesian updating mechanism, allowing nodes to dynamically adjust their decay factors based on observed graph changes. This enables real-time detection of sudden structural shifts, improving anomaly identification in streaming graphs. We evaluate Adaptive-DecayRank on multiple real-world security datasets, including DARPA and CTU-13, as well as synthetic dense graphs generated using RTM. Our experiments demonstrate that Adaptive-DecayRank outperforms state-of-the-art methods, such as AnomRank, Sedanspot, and DynAnom, achieving up to $13.94\%$ higher precision, $8.43\%$ higher AUC, and more robust detection in highly dynamic environments. Full article

For daily activity recognition in smart homes, it is possible to reduce the effort required for labeling by transferring a trained model. This involves utilizing a labeled daily activity dataset from one smart home to recognize other activities in another. The foundation of this transfer lies in establishing a shared common feature space between the two smart homes, achieved through a feature mapping approach for daily activities. However, existing heuristic feature mapping methods are often coarse, resulting in only moderate recognition performance. In this paper, we propose a fine-grained daily activity feature mapping approach. Sensors are ranked by their significance using the PageRank algorithm, and a novel alignment algorithm is introduced for sensor mapping. Experiments conducted on the publicly available CASAS dataset demonstrate that the proposed method significantly outperforms existing daily activity feature mapping approaches. Full article

Optimizing railway station operations necessitates the identification of critical track sections that constrain design throughput capacity under fixed infrastructure conditions. This paper proposes a novel multi-layer complex network-based approach for modeling and analyzing railway station yard diagrams, reframing the identification of key track sections affecting station throughput capacity as a node importance evaluation problem. In this model, nodes represent track sections included in routes specified by the station interlocking tables, while edges denote sequential connections between nodes. The structural relationships among nodes are captured using adjacency matrix (AM), structural matrix (SM), connection count matrix (CCM), and transition probability matrix (TPM). To evaluate node importance, five key indicators are introduced: connectivity strength (CS), destination node count (DNC), source node count (SNC), node efficiency (NE), and an extended PageRank (EPR). Additionally, a layered network node importance analysis method based on a single indicator, along with a comprehensive evaluation approach for the importance of the multi-layer network node, is presented. A case study conducted on a conventional railway station demonstrates that the proposed method effectively identifies key track sections through both hierarchical single-indicator evaluation and comprehensive assessment approaches. Furthermore, this paper investigates key node evaluation indicators and explores an alternative method based on Principal Component Analysis and Rank Sum Ratio (PCA-RSR), which also proves effective in identifying critical track sections. Full article

The agri-food sector is expanding, driven by growing global demand. At the same time, it faces the challenge of increasing its efficiency and adopting sustainable practices. This study aimed to map scientific production in this field, identifying trends, emerging themes, critical gaps, and future directions for research. A bibliometric analysis was conducted with 5141 papers published between 1977 and 2024, extracted from the Scopus and Web of Science databases. We applied keyword co-occurrence analysis, thematic analysis, thematic evolution, and three-field graphs using the metrics betweenness centrality, closeness centrality, and PageRank. The results revealed a significant growth in publications in the agri-food sector, especially after 2012, emphasizing the high centrality and relevance of themes such as sustainability, agri-food, and agriculture. Topics such as bioactive compounds, blockchain, and traceability were identified as areas of growing interest, and the circular economy stood out as an emerging topic. Italy, Spain, and France lead in scientific production and international collaboration. The most prominent journals were Sustainability, the Journal of Cleaner Production, and Agriculture and Human Values. Research in the sector is expanding, focusing on sustainability, the circular economy, and bioactive compounds. International collaborations and high-impact journals are pillars for advances in the sector. Full article

This article discusses current methods for ranking researchers and proposes a new metric, the output-normalized score (OnS), which considers the number of publications, citations, coauthors, and the author’s position within each publication. The proposed OnS offers a balanced approach to evaluating a researcher’s scientific contributions while addressing the limitations of widely used metrics such as the h-index and its modifications. It favors publications with fewer coauthors while giving significant weight to both the author’s position in the publication and the total number of citations. Full article

Traffic flow prediction, crucial for traffic management, relies on spatial and temporal data to achieve high accuracy. However, standard performance metrics only measure the average prediction errors and overlook the spatiotemporal aspects. To address this gap, this study introduces three simple spatial key performance indicators (KPIs): Global Moran’s I, Getis-Ord General G, and Adapted PageRank Algorithm Modified (APAM). We evaluated the traffic prediction results for synthetic clustering scenarios and four different prediction methods applied to the PeMSD8 dataset using spatial KPIs. Spatial KPIs are calculated based on traffic prediction errors and the adjacency matrix of the traffic network. Our results demonstrate that spatial KPIs can effectively differentiate between synthetic clustering scenarios. Global Moran’s I measures the spatial autocorrelation, Getis-Ord General G measures the spatial clustering of high/low values, and the univariate analysis of APAM deduces the distribution of node importance by considering node centrality and node values. Getis-Ord General G showed the highest sensitivity, being capable of distinguishing between methods with similar average RMSE, whereas Global Moran’s I and APAM univariate analysis revealed subtle differences between methods. Spatial KPIs serve as important complementary metrics for performance evaluation in the design of robust traffic management systems. Full article