Rapid classification of social media content during humanitarian crises is essential for effective disaster relief; however, traditional approaches require extensive annotated training data, which are often unavailable during new disasters. This paper presents a training-free, multimodal classification framework that leverages zero-shot vision-language models to analyze disaster-related social media content without task-specific training. The framework employs a two-stage prompt-engineered pipeline using the locally deployable Mistral-Small-3.1-24B-Instruct model, performing binary informativeness detection followed by multiclass categorization into eight humanitarian categories through structured JSON output generation. Evaluation on the CrisisMMD dataset of 18,082 multimodal samples from seven natural disasters demonstrated binary F1 scores above 0.84 for both text and image informativeness detection and weighted F1 scores of 0.61 (text) and 0.72 (image) for humanitarian categorization. The framework generalizes consistently across all disaster types with minimal performance variance (standard deviation below 0.031) and operates entirely on local infrastructure without cloud dependencies, requiring only moderate GPU resources. By eliminating training data requirements, this approach enables immediate deployment during new disasters, demonstrating that zero-shot multimodal classification achieves practically relevant performance for real-time crisis response.
This research investigates the cybersecurity awareness of young adults aged 18 to 27 through a structured survey. It evaluates key areas such as phishing recognition, two-factor authentication usage, data backup practices, and knowledge of encryption. The findings indicate that while many participants actively use two-factor authentication, there are notable gaps in understanding encryption and managing privacy settings. Differences in awareness were observed between age groups, with younger individuals displaying lower confidence in technical skills. The results highlight the importance of targeted educational programs to address these knowledge gaps and enhance online safety practices. These initiatives are essential to help young adults strengthen their cybersecurity defenses and protect their personal information in an increasingly digital world. The study underscores the continuous need for education to stay ahead of evolving cyber threats.
One major innovation in improving organizations security measures is the adoption of AI-based vulnerability scanners within the cybersecurity space. The paper analyzes crosssectional survey research identifying factors that influence the acceptance and use of such advanced tools among cybersecurity professionals. The primary method of gathering data was a structured survey questionnaire that used Likert-scale questions to quantify the participants opinions objectively. It contained 20 questions based on established models, including TAM, UTAUT, and IDT. In this research, the total number of people who responded to the survey was 49, comprising cybersecurity professionals working in various industry domains. This instrument has measured perceived usefulness, ease of use, performance expectancy, effort expectancy, social influence, facilitating conditions, and the stages of adoption, including awareness, interest, evaluation, trial, and adoption. Our results provide insight into factors that drive or hinder the adoption of AI-based vulnerability scanners, focusing on the significant role of perceived benefits and organizational support. The present paper offers valuable implications for practitioners and researchers who aim to foster AI-driven security solutions within organizational contexts.