Search Results (3,749)

Accurate identification of Chinese sensitive words is critical for maintaining online information security. However, this task faces three technical challenges: (1) high contextual dependency causing semantic ambiguity; (2) adversarial variations (e.g., homophones, character splitting) that evade exact matching; and (3) scarcity of high-quality annotated samples in complex scenarios, leading to few-shot distribution characteristics. To address these challenges, we propose a Branch Transformer Mutual-Information Calibration (BTMC) network. Specifically: (i) to capture multi-level, cross-dimensional semantic interactions despite limited data, we design a branch-based Transformer structure that aligns and fuses features across different semantic dimensions; (ii) to establish context channels between global and local semantics under few-shot conditions, we introduce a global-local interactive fusion mechanism that enhances focus on core semantics; (iii) to improve discriminability of complex semantic patterns, we propose a semantic calibration regularization mechanism that reweights features and balances information distribution. Experimental results on a newly constructed Chinese sensitive words dataset (45,623 sentences, four categories) demonstrate that BTMC achieves average F1-scores of 0.9715 (Politics and Violence), 0.9683 (Rudeness and Vulgarity), 0.9704 (Drugs and Gambling), and 0.9531 (Others), outperforming state-of-the-art baselines by 10–15% relative improvement. The code and dataset will be made publicly available. Full article

This article analyzes how twentieth- and twenty-first-century Afro-Colombian women poets from the Pacific region challenge and reframe the feminization of the land and the naturalization of the Black female body within colonial and Eurocentric epistemologies. Drawing on a framework that conceptualizes body, territory, spirituality, and community as an interdependent continuum, the article conducts close textual analysis to demonstrate how these poets construct territory and the Black female body as sentient sites. These sites are simultaneously shaped by historical violence, forced displacement, extractive economies, and racialized gender constructs, while preserving ancestral knowledge and collective memory. The findings show that Valencia Córdoba develops the body–territory through metaphor and anaphora as a generative space; Grueso Romero deploys orality and the sea as transatlantic archives of ancestry and identity; and Truque articulates urban displacement as an ontological rupture that affects memory and Black subjectivity. Ultimately, the article advances the concept of body–territory as a decolonial aesthetic and analytical tool through which Afro-Colombian women’s poetry articulates environmental justice, gendered racialization, and forms of resistance within the Afrodiasporic diaspora. Full article

This article examines how justice institutions produce and reproduce gender-based violence against women through the invisibilization of moral violence, with particular attention to their spatial dimensions. Drawing on the concept of juridical–patriarchal habitus, the study conceptualizes justice institutions not only as sites of legal action but as spatial formations that shape the visibility, recognition, and adjudication of harm. Using a feminist ethnographic approach, the article analyzes two cases of gender-based violence documented in 2020 in the municipality of Querétaro, Mexico. The findings demonstrate how movement into legal and institutional spaces transforms lived experiences of violence, as procedural requirements, evidentiary expectations, and institutional interactions operate as spatial filters that render certain forms of harm visible while obscuring others. In this process, justice actors construct and reproduce gendered stereotypes about what counts as violence, simultaneously positioning women as victims and subjecting them to processes of revictimization. By conceptualizing the invisibility of moral violence as a spatially mediated process, the article contributes to debates in legal and feminist geography, highlighting how institutional spaces not only respond to gender-based violence but actively participate in its production and concealment. Full article

Sexual violence against women frequently originates during adolescence, when structural inequalities and gendered power dynamics heighten vulnerability, making early identification of risk factors essential to prevent trajectories leading to sexual exploitation. This study presents the psychometric validation of the EDR-ESIA, a screening instrument designed to detect vulnerability to Child Sexual Exploitation (CSE) in healthcare, education, and social care settings, with particular relevance for prevention strategies targeting adolescent girls. The sample comprised 199 adolescents aged 11–17 years (M = 15.23; SD = 1.59) residing in Spain (58.8% female, 40.2% male, 1.0% unspecified), assessed by trained professionals using case records and reports. The 88-item instrument underwent expert review and pilot testing prior to validation, and its internal structure was examined using Partial Least Squares Structural Equation Modeling (PLS-SEM). The results indicated that all subdimensions and higher-order constructs showed an adequate fit to the theoretical model, supporting the instrument’s validity. Female adolescents scored significantly higher than males on CSE target indicators, reflecting a medium-to-large gender difference in vulnerability levels. Overall, the EDR-ESIA constitutes an evidence-based instrument for the timely recognition of CSE vulnerability, supporting prevention, education, and intervention efforts aimed at reducing sexual violence against women from early developmental stages. Full article

Background: Workplace victimization is a form of repeated and systematic psychological violence that can severely affect both mental and physical health. From a psychological perspective, it impacts mood states, defense mechanisms, and personality functioning. Methods: This cross-sectional study investigated the psychological and behavioural correlates of workplace victimization in a sample of 33 workers from various professional sectors, using a multidimensional assessment including standardized measures of personality traits, mood states, and defense mechanisms. Results: The MMPI-2 profile revealed elevated scores in Hypochondriasis (Hs: 72.00), Depression (D: 70.21), Hysteria (Hy: 67.61), and Paranoia (Pa: 68.76), indicating somatic symptoms, depressive features, and suspiciousness. The POMS showed increased Tension–Anxiety (T: 65.06), Depression–Dejection (D: 68.21), Anger–Hostility (A: 68.15), and Fatigue–Inertia (F: 65.24), alongside reduced Vigor–Activity (V: 43.18). The DMI analysis highlighted a high Reversal score (REV: 65.91), suggesting a predominant use of defense mechanisms such as altruism and idealization to cope with distress. Conclusions: In this selected sample of adults referred for psychological evaluation for suspected or documented workplace victimization, participants showed a clinically relevant psychological burden, including depressive symptoms, somatic concerns, Anger–Hostility, fatigue, reduced vigor, and specific defensive patterns. Given the cross-sectional design, small sample size, and absence of a control group, these findings should be interpreted as preliminary and cannot establish causality or the specificity of this profile to workplace victimization. Full article

Mistreatment is a pervasive and normalized feature of medical culture. In medical residencies, it functions as a structural rite of passage that shapes professional socialization. While the prevalence of mistreatment is documented, there is a lack of qualitative research exploring its role as a mechanism of identity construction. The aim of this study was to understand the experiences of mistreatment among internal medicine residents in Medellín, Colombia, through the lens of ritual theory and symbolic violence. A particularistic ethnographic study was conducted with 12 residents selected via theoretical sampling. Data were collected through semi-structured interviews and a reflexive field journal. Rigor was ensured using investigator triangulation and analytical bracketing to manage researchers’ biases. The training process follows a three-stage rite. (1) Separation: Symbolic violence and social pressure to specialize frame general medicine as “mediocre,” turning admission into a “battlefield” where self-worth is tied to success. (2) Marginalization (Liminality): Residents endure systemic mistreatment, including sleep deprivation (3.5 h rest cycles), public ridicule (“pimping”), and physical/verbal abuse (e.g., being hit with stethoscopes or called “testicles/jerks”). This stage is governed by a “purificatory logic” where suffering is internalized as a meritocratic requirement. This leads to high morbidity, with clinical diagnoses of anxiety and depression. (3) Integration (Postliminality): Professional autonomy and financial stability act as a “redemption” that justifies past suffering. Mistreatment is not an isolated interpersonal issue but a structurally embedded ritual and a core element of the hidden curriculum. It reinforces toxic hierarchies and a “tyranny of merit” that obscures structural barriers. These findings offer analytically transferable insights for global medical education, calling for a deconstruction of ritualized violence to foster more humanistic training environments. Full article

This concept paper advances stigma power as a central analytical mechanism for understanding how patriarchy, capitalism, white supremacy, and cis-heteronormativity operate with particular intensity against sex workers. Integrating Link and Phelan’s stigma power with Bourdieu’s symbolic violence and Foucauldian productive power, the framework theorises stigma as a mechanism institutionalised through law and enforced by institutions, which produces measurable consequences that include violence, exclusion, and health harms. Analysing the intersecting axes of gender, sexuality, race, migration, and class across three qualitative studies (SWMH, SEXHUM, VICSW), the article demonstrates why labour-rights reforms, including decriminalisation, are necessary but insufficient. Dismantling stigma requires not only removing sanctions but actively contesting the actors exercising stigma power and interrupting the stabilising mechanisms that reproduce it. This requires policy that acknowledges stigma’s existence whilst working to dismantle it, rather than eliding its reality through liberal mainstreaming or strengthening it through criminalisation or rescue frameworks. The framework explains why decriminalisation is associated with better access to rights and health; why all criminalisation including the so-called Swedish model correlates with increased violence; why stigma persists under optimal legal conditions; and how intersecting marginalisations produce differential vulnerability. Policy implications emphasise pairing decriminalisation with peer-led anti-stigma work, institutional reform, migrant rights, and funded support for sex worker self-organisation. Full article

Domestic Violence affects 1 in 3 women worldwide. Empirical evidence from India suggests that women and girls experience a continuum of violence and discrimination from prenatal stages till death in families that largely continue to operate within a dominantly patriarchal framework. However, the literature on domestic violence in India suffers from problems pertaining to reductive and episodic framing, focusing on short-term prevalence, and frames the impact on survivors largely in terms of clinical constructs such as anxiety, depression, and PTSD. This work argues for a broader, thematic framing of domestic and familial violence and contends that the psychological sequelae of this kind of chronic and systemic discrimination and violence cannot be captured using rigid clinical constructs that dominate psychological literature. We propose a conceptual framework, i.e., SPECTRA (Socially and Psychologically Embedded Continuous Trauma in Relational Architecture), which is partially aligned with the propositions of complex trauma. However, we also critique the origin of complex trauma within hegemonic psychiatry and highlight the need for creating a culturally adapted expansion—to shift the emphasis from an individually rooted, diagnostic framework to a culturally contextualized continuous trauma framework. We utilize seven illustrative case studies to define the tenets of the SPECTRA model. Full article

Intimate partner violence (IPV) is a significant global public health issue that requires accessible, scalable, and contextually appropriate interventions. Mobile health (mHealth) technologies provide a promising platform to deliver support, information, and safety planning tools for individuals at risk of IPV. This study aimed to develop and pilot-test iCanPlan, a mobile application designed to support IPV prevention in Thailand. The application evaluates IPV risk, identifies indicators of danger, and provides a countrywide list of assistance sources. iCanPlan consists of four main components: (1) an IPV risk assessment tool, (2) a list of support resources, (3) educational materials presented in the form of infographics, and (4) encouraging quotes from well-known public figures. The app features a clean, user-friendly interface with intuitive navigation and color-coded components to enhance usability. In addition, a preliminary study was conducted with 30 experts from multidisciplinary fields, including gender-based violence research, social work, psychology, public health, and non-governmental organizations. Participants used the application for one month and subsequently evaluated it using a structured questionnaire based on heuristic evaluation principles. The questionnaire assessed usability, safety features, content quality, cultural appropriateness, language clarity, ethical considerations, and overall evaluation using a five-point Likert scale. Data was analyzed using descriptive statistics (mean and standard deviation) in SPSS. The findings demonstrated excellent performance across all domains, with high mean scores for usability (M = 4.93), safety features (M = 4.73), and content quality (M = 4.82), while cultural appropriateness, language clarity, ethical considerations, and overall evaluation achieved perfect scores (M = 5.00). These results indicate strong agreement among experts regarding the application’s usability, safety, and relevance. The study highlights the potential of iCanPlan as a culturally appropriate and user-friendly digital intervention for IPV prevention. Further research involving the target population is needed to evaluate its effectiveness and long-term impact on help-seeking behavior and IPV-related outcomes. Full article

Accurate and consistent forensic bruise assessment is critical in ensuring positive clinical and legal outcomes for victims of violence. In this study, a framework for automated bruise detection is presented that, for the first time, integrates narrowband alternate-light-source (ALS) forensic imaging and ambient white light imaging. This evaluation framework is designed to address long-standing issues with respect to equitable performance across skin tones and lighting scenarios via a combination of novel model diagnostic strategies. In particular, skin-tone balancing during training and testing, threshold-sensitivity analysis, and embedding-similarity partitioning are employed to quantify the model robustness and deployment trade-offs that arise in forensic image analysis. Models were implemented with ImageNet-pretrained backbones and trained on a unique, multi-annotator full-consensus dataset comprising both white-light and ALS (415 nm and 450 nm) images. The protocol emphasizes three axes of operational relevance: (1) illumination composition in training (W/ALS ratio); (2) subgroup fairness via targeted balancing; and (3) model operating-point selection (confidence and IoU thresholds) informed by confidence-stability metrics and bootstrapped uncertainty estimates. Systematic W/ALS ratio sweeps indicate peak accuracy under ALS-dominant training and declining performance as the proportion of white-light images increases within the training set. Skin-tone balancing reduced failure rates for darker skin tones but increased overprediction in some demographic subgroups. Embedding-similarity and seen/unseen injury analyses demonstrate inflated generalization under image-level partitioning. Ultimately, the findings suggest that future researchers and developers should employ injury-level data partitioning and ensure a weighted balance of ALS images during training. Full article

Background/Objectives: Despite legal progress and achievements regarding trans* rights in Portugal over recent decades, trans* individuals still face high levels of violence and discrimination, especially within family and intimate relationships. This qualitative study aimed to explore the experiences and perspectives of professionals working in the domestic violence field with the trans* population and service provision within the National Support Network for Victims of Domestic Violence (RNAVVD). Methods: Semi-structured interviews were conducted with eight participants, including psychologists, social workers, and program directors from organizations supporting victims of domestic violence in Portugal. Data were analyzed using Codebook Thematic Analysis to identify themes, resulting in two main themes: (1) Experiences in Working with Victimized Trans* Individuals; and (2) Framework of Portugal’s National Support Network. Results: The results showed that trans* individuals face significant vulnerabilities due to family and intimate partner violence, systemic discrimination, and inequalities in essential services. Young trans* individuals are seen as being particularly at risk due to the impacts of violence and lack of family support on their autonomy, and additional barriers to entering the labor market. Participants also identified barriers faced by this population when trying to access victim support services (e.g., lack of specialized training and low availability of specialized and culturally competent services), while highlighting efforts by LGBTQIA+ services to meet their psychosocial needs. Conclusions: In conclusion, public institutions must address the specific needs of trans* individuals by developing policies and services that adopt a cross-sectoral, intersectional approach across society. Full article

Background: The aim of this study was to examine the pattern of child-to-parent violence (CPV) across a broad age range, from early adolescence to late emerging adulthood. Specifically, the objectives were to analyze the linear and quadratic relationships between CPV types (psychological, physical, financial, and control/domain behaviors) and age, as well as to examine the interaction of sex within this relationship. Methods: A total of 1959 adolescents (13–17 years) and 1046 young adults (18–25 years) completed, respectively, the adolescent and young adult versions of the Child-to-Parent Violence Questionnaire (CPV-Q). Results: Age was curvilinearly associated with psychological CPV (increasing until approximately age 19 and then decreasing), positively linearly associated with financial CPV (increasing with age), and negatively linearly associated with control/domain behaviors (decreasing with age). No significant association was found between age and physical CPV. Furthermore, boys and girls showed different age-related patterns in some CPV types. Conclusions: These findings suggest that CPV does not disappear after adolescence, and that the pattern is not uniform throughout development nor the same for boys and girls. The results (1) underscore the importance of studying CPV considering developmental stage, sex, and the specific CPV types, and (2) may contribute to facilitate the early detection of CPV, anticipating changes in violence patterns, and guiding prevention strategies tailored to each developmental stage. Full article

Children increasingly consume online video content, creating a growing need for scalable approaches to support content moderation workflows. However, directly identifying harmful or policy-violating content, such as violence, sexual content, or self-harm, remains a complex task that typically requires specialized classifiers and domain-specific annotations. In this context, sentiment analysis can provide complementary information by capturing affective signals expressed through language and visual cues. This study does not treat sentiment polarity as a direct indicator of unsafe or policy-violating content. Instead, it explores multimodal sentiment analysis as an auxiliary triage signal that may help prioritize content for human review or identify segments requiring further inspection. This paper investigates the feasibility of using large vision–language models (LVLMs) for zero-shot multimodal sentiment analysis on utterance-aligned video segments. We evaluate two LVLMs, LLaVA-OneVision-7B and Qwen2.5-VL-7B, under three input settings: text-only, vision-only, and multimodal, using a conversational TV-series dataset consisting of short utterance-level video segments and transcripts. The results show that multimodal sentiment inference can provide useful screening signals without task-specific fine-tuning, although the benefits are model-dependent. LLaVA-OneVision-7B consistently outperforms Qwen2.5-VL-7B and benefits more clearly from combining textual and visual inputs, whereas Qwen2.5-VL-7B shows limited improvement across modality settings. We also analyze the trade-off between frame sampling and image resolution. Finally, we discuss limitations related to dataset scope, annotation subjectivity, class imbalance, and the need for broader validation before real-world deployment. Full article

Alcohol and psychoactive substance use represent a major burden for global public health, increasing the risk of non-communicable diseases, violence, road traffic injuries, dependence, and mental disorders, and generating impacts on productivity and social welfare. This study aimed to estimate the burden of disease attributable to alcohol and other psychoactive substances in the departments of Colombia from 2016 to 2022. A burden-of-disease study was conducted using the Disability-Adjusted Life Years (DALYs) indicator, following the methodology of the World Health Organization Global Health Estimates. Official morbidity and mortality databases were used. An estimated 236,154.42 DALYs were attributable to alcohol and psychoactive substance use in Colombia during the study period, increasing from 14,158.7 DALYs in 2016 to 40,190.7 DALYs in 2022. The burden was heterogeneous across departments, with values above 1000 DALYs in Quindío (1779.5), Nariño (1624.3), and Norte de Santander (1008.0) and below 132 DALYs in La Guajira, Casanare, and Vaupés. Men accounted for 73.5% of total DALYs. The mean age of morbidity records associated with alcohol and psychoactive substance use disorders was 30.67 years in men and 32.37 years in women. The burden associated with psychoactive substance use is increasing in Colombia, with differences by sex and department of residence. Full article

Background: Violence in healthcare settings affects nursing students during clinical training and may compromise their mental well-being, learning experiences, and professional development. Despite evidence from high-income countries, limited data exist on how contextual and organizational factors in Latin American settings shape these experiences. This study aimed to assess the frequency and types of violence experienced by nursing students during clinical practice and the academic and emotional consequences. Methods: Cross-sectional study conducted among undergraduate and graduate nursing students in Mexico City who had completed at least one hospital-based clinical placement in the previous 12 months. Data were collected between January 2024 and September 2025 using a validated questionnaire assessing types of violence, perpetrators, academic and emotional consequences. Violence was defined as experiencing events “occasionally,” “sometimes,” or “frequently.” Descriptive statistics were calculated. Associations were examined using Pearson’s chi-square test, and logistic regression models adjusted for age, sex, and year of study. Results: Seventy-three students participated (86.3% female). Non-physical violence was the most frequent type (90.4%), followed by sexual harassment (49.3%), mainly perpetrated by nurses (62%) and physicians (46.5%). Considering leaving the profession (41.4%) and feelings of inadequacy (66.2%) were the most common academic and emotional consequences. Although some associations were observed in bivariate analyses, these were not significant after adjustment. Conclusions: Violence during clinical training is highly prevalent and may represent a significant threat to nursing students’ mental well-being and professional development. Strengthening institutional policies, reporting mechanisms, and supportive learning environments is essential to mitigate its impact and promote safer clinical training. Full article