Search Results (282)

In cybersecurity, constructing an accurate knowledge graph is vital for discovering key entities and relationships in security incidents buried in vast unstructured threat reports. Traditional knowledge-graph construction pipelines based on handcrafted rules or conventional machine learning models falter when the data scale and linguistic variety grow. GraphRAG, a retrieval-augmented generation (RAG) framework that splits documents into fixed-length chunks and then retrieves the most relevant ones for generation, offers a scalable alternative yet still suffers from fragmentation and semantic gaps that erode graph integrity. To resolve these issues, this paper proposes SC-LKM, a cybersecurity knowledge-graph construction method that couples the GraphRAG backbone with hierarchical semantic chunking. SC-LKM applies semantic chunking to build a cybersecurity knowledge graph that avoids the fragmentation and inconsistency seen in prior work. The semantic chunking method first respects the native document hierarchy and then refines boundaries with topic similarity and named-entity continuity, maintaining logical coherence while limiting information loss during the fine-grained processing of unstructured text. SC-LKM further integrates the semantic comprehension capacity of Qwen2.5-14B-Instruct, markedly boosting extraction accuracy and reasoning quality. Experimental results show that SC-LKM surpasses baseline systems in entity-recognition coverage, topology density, and semantic consistency. Full article

In the era of homogenized competition, brand knowledge has become a critical factor that influences consumer purchasing decisions. However, traditional single-layer network models fail to capture the multi-dimensional semantic relationships embedded in brand-related textual data. To address this gap, this study proposes a BKMN framework integrating TF-IDF and TextRank algorithms for comprehensive brand knowledge discovery. By analyzing 19,875 consumer reviews of a mobile phone brand from JD website, we constructed a tri-layer network comprising TF-IDF-derived keywords, TextRank-derived keywords, and their overlapping nodes. The model incorporates co-occurrence matrices and centrality metrics (degree, closeness, betweenness, eigenvector) to identify semantic hubs and interlayer associations. The results reveal that consumers prioritize attributes such as “camera performance”, “operational speed”, “screen quality”, and “battery life”. Notably, the overlap layer exhibits the highest node centrality, indicating convergent consumer focus across algorithms. The network demonstrates small-world characteristics (average path length = 1.627) with strong clustering (average clustering coefficient = 0.848), reflecting cohesive consumer discourse around key features. Meanwhile, this study proposes the Mul-LSTM model for sentiment analysis of reviews, achieving a 93% sentiment classification accuracy, revealing that consumers have a higher proportion of positive attitudes towards the brand’s cell phones, which provides a quantitative basis for enterprises to understand users’ emotional tendencies and optimize brand word-of-mouth management. This research advances brand knowledge modeling by synergizing heterogeneous algorithms and multilayer network analysis. Its practical implications include enabling enterprises to pinpoint competitive differentiators and optimize marketing strategies. Future work could extend the framework to incorporate sentiment dynamics and cross-domain applications in smart home or cosmetic industries. Full article

Modeling accident investigation reports is crucial for elucidating accident causation mechanisms, analyzing risk evolution processes, and formulating effective accident prevention strategies. However, such reports are typically long, hierarchically structured, and information-dense, posing unique challenges for existing language models. To address these domain-specific characteristics, this study proposes SAFE-Transformer, a Structure-Aware and Format-Enhanced Transformer designed for long-document modeling in the emergency safety context. SAFE-Transformer adopts a dual-stream encoding architecture to separately model symbolic section features and heading text, integrates hierarchical depth and format types into positional encodings, and introduces a dynamic gating unit to adaptively fuse headings with paragraph semantics. We evaluate the model on a multi-label accident intelligence classification task using a real-world corpus of 1632 official reports from high-risk industries. Results demonstrate that SAFE-Transformer effectively captures hierarchical semantic structure and outperforms strong long-text baselines. Further analysis reveals an inverted U-shaped performance trend across varying report lengths and highlights the role of attention sparsity and label distribution in long-text modeling. This work offers a practical solution for structurally complex safety documents and provides methodological insights for downstream applications in safety supervision and risk analysis. Full article

Introduction: Huntington’s disease (HD) disrupts cortico-striato-thalamocortical circuits decades before clinical onset. Electroencephalography (EEG) offers millisecond temporal resolution, low cost, and broad accessibility, yet its mechanistic and biomarker potential in HD remains underexplored. We conducted a mechanistic review to synthesize half a century of EEG findings, identify reproducible electrophysiological signatures, and outline translational next steps. Methods: Two independent reviewers searched PubMed, Scopus, Google Scholar, ResearchGate, and the Cochrane Library (January 1970–April 2025) using the terms “EEG” OR “electroencephalography” AND “Huntington’s disease”. Clinical trials published in English that reported raw EEG (not ERP-only) in human HD gene carriers were eligible. Abstract/title screening, full-text appraisal, and cross-reference mining yielded 22 studies (~700 HD recordings, ~600 controls). We extracted sample characteristics, acquisition protocols, spectral/connectivity metrics, and neuroclinical correlations. Results: Across diverse platforms, a consistent spectral trajectory emerged: (i) presymptomatic carriers show a focal 7–9 Hz (low-alpha) power loss that scales with CAG repeat length; (ii) early-manifest patients exhibit widespread alpha attenuation, delta–theta excess, and a flattened anterior-posterior gradient; (iii) advanced disease is characterized by global slow-wave dominance and low-voltage tracings. Source-resolved studies reveal early alpha hypocoherence and progressive delta/high-beta hypersynchrony, microstate shifts (A/B ↑, C/D ↓), and rising omega complexity. These electrophysiological changes correlate with motor burden, cognitive slowing, sleep fragmentation, and neurovascular uncoupling, and achieve 80–90% diagnostic accuracy in shallow machine-learning pipelines. Conclusions: EEG offers a coherent, stage-sensitive window on HD pathophysiology—from early thalamocortical disinhibition to late network fragmentation—and fulfills key biomarker criteria. Translation now depends on large, longitudinal, multi-center cohorts with harmonized high-density protocols, rigorous artifact control, and linkage to clinical milestones. Such infrastructure will enable the qualification of alpha-band restoration, delta-band hypersynchrony, and neurovascular coupling as pharmacodynamic readouts, fostering precision monitoring and network-targeted therapy in Huntington’s disease. Full article

Background/Objectives: Dyslexia, a learning disability affecting reading, has been extensively studied using eye movements. This study aimed to examine in the same design the effects of different psycholinguistic variables, i.e., grammatical category, lexical frequency, word length and orthographic consistency on eye movement patterns during reading in adults. Methods: We compared the eye movements of forty university students, twenty with and twenty without dyslexia while they read aloud a meaningful and a meaningless text in order to examine whether semantic context could enhance their reading strategy. Results: Dyslexic participants made more reading errors and had longer reading time particularly with the meaningless text, suggesting an increased reliance on the semantic context to enhance their reading strategy. They also made more progressive and regressive fixations while reading the two texts. Similar results were found when examining grammatical categories. These findings suggest a reduced visuo-attentional span and reliance on a serial decoding approach during reading, likely based on grapheme-to-phoneme conversion. Furthermore, in the whole text analysis, there was no difference in fixation duration between the groups. However, when examining word length, only the control group exhibited a distinction between longer and shorter words. No significant group differences emerged for word frequency. Importantly, multiple regression analyses revealed that orthographic consistency predicted fixation durations only in the control group, suggesting that dyslexic readers were less sensitive to phonological regularities—possibly due to underlying phonological deficits. Conclusions: These findings suggest the involvement of both phonological and visuo-attentional deficits in dyslexia. Combined remediation strategies may enhance dyslexic individuals’ performance in phonological and visuo-attentional tasks. Full article

Readability formulas are mathematical functions that assess the ‘difficulty’ level of a given text. They play a crucial role in aligning educational texts with student reading abilities; however, existing models are often not tailored to specific linguistic or regional contexts. This study aims to develop and evaluate two novel readability formulas specifically designed for the Mexican Spanish language, targeting elementary education levels. The formulas were trained on a corpus of 540 texts drawn from official elementary-level textbooks issued by the Mexican public education system. The first formula was constructed using multiple linear regression, emulating the structure of traditional readability models. The second was derived through genetic programming (GP), a machine learning technique that evolves symbolic expressions based on training data. Both approaches prioritize interpretability and use standard textual features, such as sentence length, word length, and lexical and syntactic complexity. Experimental results show that the proposed formulas outperform several well-established Spanish and non-Spanish readability formulas in distinguishing between grade levels, particularly for early and intermediate stages of elementary education. The GP-based formula achieved the highest alignment with target grade levels while maintaining a clear analytical form. These findings underscore the potential of combining machine learning with interpretable modeling techniques and highlight the importance of linguistic and curricular adaptation in readability assessment tools. Full article

In this paper, we investigate the approximate string matching problem when the allowed edit operations are non-overlapping unbalanced translocations of adjacent factors. This kind of edit operation takes place when two adjacent substrings of the text swap, resulting in a modified string. The two involved substrings are allowed to be of different lengths. Such large-scale modifications of strings have various applications, notably in fields such as computational biology and genomics, where structural rearrangements play a key role. However, despite their central role in many fields of text processing, little attention has been devoted to the problem of matching strings allowing for this kind of edit operation. In this paper, we present three algorithms for solving the problem, all of them with an $\mathcal{O}\left(n{m}^3\right)$ worst-case and an $\mathcal{O}\left(m^2\right)$-space complexity, where m and n are the length of the pattern and of the text, respectively. Specifically, our first algorithm is based on the dynamic programming approach. Our second solution improves the previous one by making use of the Directed Acyclic Word Graph of the pattern. Finally, our third algorithm is based on an alignment procedure. We also show that under the assumptions of equiprobability and independence of characters, our second algorithm has an $\mathcal{O}\left(n{log}_{\sigma }^{2}m\right)$ average time complexity for an alphabet of size $\sigma \ge 4$. Full article

Background: The UK Department for Digital, Culture, Media, and Sport (DCMS) grassroots concussion guidance, May 2023, advised that all community-based sport-related concussions (SRCs) be diagnosed by a healthcare practitioner. This may require that general practitioners (GPs) diagnose and manage SRCs. Diagnosing SRCs in primary care settings in the United Kingdom (UK) presents significant challenges, primarily due to the lack of validated tools specifically designed for general practitioners (GPs). This scoping review aims to identify diagnostic and management tools for SRCs in grassroots sports and primary care settings. Aims: To identify tools that can be used by GPs to diagnose and manage concussions in primary care, both adult and paediatric populations. Design and Methods: A scoping review was conducted in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses extension for Scoping Reviews (PRISMA-ScRs). Five databases (MEDLINE, EMBASE, CINAHL, Cochrane Library, Google Scholar) were searched from 1946 to April 2025. Search terms included “concussion”, “primary care”, and “diagnosis”. Studies that discussed SRCs in community or primary care settings were included. Those that exclusively discussed secondary care and elite sports were excluded, as well as non-English studies. Two reviewers independently screened titles, abstracts, and full texts, with a third resolving any disagreements. Data were extracted into Microsoft Excel. Studies were assessed for quality using the Joanna Briggs critical appraisal tools and AGREE II checklist. Results: Of 727 studies, 12 met the inclusion criteria. Identified tools included Sport Concussion Assessment Tool 6 (SCAT6, 10–15 min, adolescent/adults), Sport Concussion Office Assessment Tool 6 (SCOAT6, 45–60 min, multidisciplinary), the Buffalo Concussion Physical Examination (BCPE, 5–6 min, adolescent-focused), and the Brain Injury Screening Tool (BIST, 6 min, ages 8+). As part of BCPE, a separate Telehealth version was developed for remote consultations. SCAT6 and SCOAT6 are designed for healthcare professionals, including GPs, but require additional training and time beyond typical UK consultation lengths (9.2 min). BIST and BCPE show promise but require UK validation. Conclusions: SCAT6, SCOAT6, BIST, and BCPE could enhance SRC care, but their feasibility in UK primary care requires adaptation (e.g., integration with GP IT systems and alignment with NICE guidelines). Further research is required to validate these tools and assess additional training needs. Full article

Background: The occurrence (incidence or prevalence) of delirium in brain tumors is unknown, yet delirium is associated with increased morbidity and mortality and worse quality of life. We conducted a systematic review and meta-analysis to determine the occurrence of delirium in hospitalized patients with brain tumors. Methods: PubMed, Scopus, and Web of Science were systematically searched for papers from 1 January 1999 to 12 July 2024, including references from texts. Cross-sectional, prospective, and other cohort study designs were included, and individual case reports, case series, editorials, and reviews were excluded. The included papers were scored using a validated sensitivity analysis tool and tested for quality and bias using funnel plots and Egger’s test. We used random effects models for the summary estimates. We performed subgroup analyses by tumor type, tumor location, delirium subtype, and length of stay. Results: Of the 452 studies screened, 27 were included, representing 35,958 patients. The overall occurrence of delirium was 0.17 (95% CI [0.11–0.24]). Delirium occurrence in patients with low-grade gliomas, high-grade gliomas, and brain metastases was 0.10 [0.06–0.16], 0.21 [0.10–0.40], and 0.31 [0.16–0.50], respectively. Compared to the occipital lobe, there was a higher occurrence of delirium for tumors in the frontal (RR 3.08 [1.35–8.22]) and temporal lobes (RR 2.88 [1.22–7.93]). The patients were more likely to have hypoactive (RR 1.61 [1.30; 1.98]) than hyperactive delirium. Delirium was associated with 4.62 additional hospitalized days compared to those without delirium (CI [3.23–6.01]). Discussion: We confirmed high occurrence rates of delirium in patients hospitalized with brain tumors. Patients with brain metastases had a higher occurrence of delirium compared to patients with gliomas, and delirium occurrence rates were higher in patients with frontotemporal tumors. Delirium occurrence rates in the literature are very heterogeneous and point toward a need for tailored assessments in patients with brain tumors. Full article

The style change detection (SCD) task asks to find the positions of authors’ style changes within multi-authored texts. It has several application areas, such as forensics, cybercrime, and literary analysis. Since 2017, SCD solutions in English have been actively investigated. However, to the best of our knowledge, this task has not yet been investigated in Arabic text. Moreover, most existing SCD solutions represent boundaries surrounding segments by concatenating them. This shallow concatenation may lose style patterns within each segment and also increase input lengths while several embedding models restrict these lengths. This study seeks to bridge these gaps by introducing an Edge Convolutional Neural Network for the Arabic SCD task (ECNN-ASCD) solution. It represents boundaries as standalone learnable parameters across layers based on graph neural networks. ECNN-ASCD was trained on an Arabic dataset containing three classes of instances according to difficulty level: easy, medium, and hard. The results show that ECNN-ASCD achieved a high $F_1$ score of 0.9945%, 0.9381%, and 0.9120% on easy, medium, and hard instances, respectively. The ablation experiments demonstrated the effectiveness of ECNN-ASCD components. As the first publicly available solution for Arabic SCD, ECNN-ASCD would open the door for more active research on solving this task and contribute to boosting research in Arabic NLP. Full article

Product type plays a critical role in shaping how consumers generate, perceive, and utilize online reviews in decision-making. While previous studies have examined various review features, this study highlights the distinct effects of product classification—search goods, experience goods, and credence goods—on both review generation and perceived helpfulness. Drawing on Product Classification Theory, as well as Self-Determination Theory and the Theory of Planned Behavior, we analyze how review characteristics such as text length and photo inclusion vary across product types and influence consumer perceptions. Using a large-scale dataset of verified Amazon reviews, we find that consumers are more likely to produce longer and more visually rich reviews for search and experience goods than for credence goods, which are harder to evaluate and, thus, elicit less elaborate content. In terms of review helpfulness, reviews for experience goods are rated as more helpful than those for credence goods, while those for search goods are seen as less helpful. Furthermore, review length significantly boosts helpfulness for search goods, while photo inclusion enhances it for experience goods. These findings contribute to review effectiveness research by emphasizing the moderating role of product type and offering actionable insights for e-commerce platforms to improve review design and consumer decision-making. Full article

Large Language Models (LLMs) have shown a strong performance across various tasks but still face challenges in automatic text summarization. While they are effective in capturing semantic patterns from large corpora, they typically lack mechanisms for encoding structural relationships between sentences or paragraphs. Their high hardware requirements and limited analysis as to processing efficiency further constrain their applicability. This paper proposes a framework employing the Graph Independent Set approach to extract the essence of textual graphs and address the limitations of LLMs. The framework encapsulates nodes and relations into structural graphs generated through Natural Language Processing (NLP) techniques based on the Maximum Independent Set (MIS) theory. The incorporation of graph-derived structural features enables more semantically cohesive and accurate summarization outcomes. Experiments on the Document Understanding Conference (DUC) and Cable News Network (CNN)/DailyMail datasets are conducted with different summary lengths to evaluate the performance of the framework. The proposed method provides up to a 41.05% (Recall-Oriented Understudy for Gisting Evaluation, ROUGE-2 F1) increase in summary quality and a 60.71% improvement in response times on models such as XLNet, Pegasus, and DistilBERT. The proposed framework enables more informative and concise summaries by embedding structural relationships into LLM-driven semantic representations, while reducing computational costs. In this study, we explore whether integrating MIS-based graph filtering with LLMs significantly enhances both the accuracy and efficiency of extractive text summarization. Full article

Purpose: The purpose of this scoping review was to systematically identify and summarize the existing literature on non-spinal clinical applications of EOS imaging and identify related evidence gaps. Method: The study followed the PRISMA-ScR guidelines. A systematic literature search was conducted in Embase, MEDLINE, CINAHL, Scopus, Cochrane, Academic Search Premier, and OpenGrey databases in November 2022 and updated in December 2023. Original research from 2003 to 2023 was eligible if in English, Danish, French, German, Norwegian, or Swedish. Two authors screened articles by title and abstract, while data extraction from full texts was performed by seven authors using a structured template. Results: A total of 8176 articles were identified, with 1350 selected for full-text review and 268 included in data extraction. Among adults, 187 articles were included, with 88 focused on surgical applications like hip arthroplasty or osteotomy. In pediatrics, 68 general and 13 surgery-related articles were included. Lower extremity analysis was the most frequent topic, with other uses identified, such as rib cage geometry, patellar dislocation, and X-linked hypophosphatemia. Conclusions: Key clinical applications of EOS imaging include lower extremity analysis, e.g., leg length assessment and knee/hip arthroplasty planning), pelvic and spinal alignment studies, and emerging uses in rib cage geometry. Evidence gaps include limited research on the diagnostic accuracy of EOS for cerebral shunt placement, reliability in bone age estimation, and an unclear role in foot and ankle morphology. Full article

In this article, we tested some specific claims made in the literature on relative distances among European languages and among Chinese dialects, suggesting that some language varieties within the Sinitic family traditionally called dialects are, in fact, more linguistically distant from one another than some European varieties that are traditionally called languages. More generally, we examined whether distances among varieties within and across European language families were larger than those within and across Sinitic language varieties. To this end, we computed lexico-phonetic as well as syntactic distance measures for comparable language materials in six Germanic, five Romance and six Slavic languages, as well as for six Mandarin and nine non-Mandarin (‘southern’) Chinese varieties. Lexico-phonetic distances were expressed as the length-normalized MPI-weighted Levenshtein distances computed on the 100 most frequently used nouns in the 32 language varieties. Syntactic distance was implemented as the (complement of) the Pearson correlation coefficient found for the PoS trigram frequencies established for a parallel corpus of the same four texts translated into each of the 32 languages. The lexico-phonetic distances proved to be relatively large and of approximately equal magnitude in the Germanic, Slavic and non-Mandarin Chinese language varieties. However, the lexico-phonetic distances among the Romance and Mandarin languages were considerably smaller, but of similar magnitude. Cantonese (Guangzhou dialect) was lexico-phonetically as distant from Standard Mandarin (Beijing dialect) as European language pairs such as Portuguese–Italian, Portuguese–Romanian and Dutch–German. Syntactically, however, the differences among the Sinitic varieties were about ten times smaller than the differences among the European languages, both within and across the families—which provides some justification for the Chinese tradition of calling the Sinitic varieties dialects of the same language. Full article

Hamstring strain injuries (HSIs) are the most common time loss injury sustained in male Australian Football League (AFL) athletes, causing significant financial cost, time cost, and impaired team and individual performance. In a squad of 42 players, HSIs accounted for 4.86 new injuries sustained by players per club per AFL season in 2020. This is consistent with injury reporting over the last decade in AFL, despite best efforts to reduce the rate. This scoping review sought to firstly identify the reported hamstring injury prevention risk factors in elite AFL, discern the impact of these factors, and map the gaps in the current literature using a biopsychosocial understanding of injury prevention. The scoping review process was based on the Askey and O’Malley framework. Five relevant online databases (MEDLINE, Proquest, CINAHL, SPORTdiscuss, and EMBASE) were systematically searched using a series of Boolean and operator terms following the PRISMA-ScR protocol using the criteria: (1) assessing male professional/elite athletes in AFL; (2) written in English and peer-reviewed; (3) full text available; and (4) published after 2006. Only manuscripts that fit the search terms and inclusion criteria were retained in the scoping review. Following an initial search, 246 potential studies were identified, with 12 studies meeting the inclusion criteria after full-text screening. The risk factors examined were subclassified into modifiable and non-modifiable categories. Modifiable factors include high-speed running exposure, gluteus medius activation, eccentric hamstring strength, shorter bicep femoris fascicle length, use of interchange, and hamstring stiffness. Non-modifiable factors include previous history of HSI and limb injury, age, and size of injury on MRI. This scoping review highlights the need for continued monitoring of high-speed running volumes as rapid increases in completed distances present as a substantial risk factor. The modifiable mechanistic risk factors of eccentric hamstring strength and hamstring stiffness were identified as important components of player screening to reduce the risk of future HSI. Risk factors identified throughout will help develop comprehensive injury profiling for athletes. Further research is warranted to develop a holistic approach to injury profiling. Full article