International Journal of Knowledge Processing Studies

Abstract The increasing frequency of road accidents highlights the urgent need for intelligent systems capable of real-time incident detection to support rapid emergency response. This paper presents a lightweight spatiotemporal accident detection framework that integrates a single-stage object detection model with an auxiliary motion-analysis stream based on dense optical flow. By jointly exploiting spatial appearance information and short-term temporal motion variations, the proposed system aims to improve robustness in complex traffic scenes while maintaining real-time performance. Unlike conventional approaches that rely solely on frame-wise object detection, the framework captures motion irregularities surrounding collision events to mitigate false alarms caused by normal traffic dynamics. A multi-source dataset was curated from publicly available traffic surveillance images and accident-related video clips obtained from heterogeneous sources, encompassing diverse viewpoints, traffic densities, and environmental conditions. The system was evaluated against representative object detection baselines using standard detection metrics, along with inference time analysis to assess deployment feasibility. Experimental results demonstrate that the proposed fusion-based approach achieves improved detection consistency with low computational overhead, making it suitable for real-time surveillance applications. The study highlights the effectiveness of combining spatial detection with simple temporal motion cues for practical accident monitoring in intelligent transportation systems, while also discussing current limitations and directions for future enhancement.

Abstract Purpose
This study investigates the structure and dynamics of entrepreneurial discourse on the Arabic Twitter (X) sphere within the Gulf Cooperation Council (GCC) states. It examines how entrepreneurship is socially constructed amidst the transition from a rentier to a knowledge-based economy, focusing on state narratives versus public sentiment.
Design/methodology/approach
Adopting a computational social science approach, the study analyzed a corpus of 48,841 content units harvested between June and December 2025. To ensure statistical independence and prevent double-voting bias, the research employed a Confidence-Calibrated Ensemble architecture. This pipeline integrates fine-tuned models (MARBERT) with pseudo-independent Large Language Model configurations (GPT-4 in zero-shot and few-shot settings) using calibrated tie-breaking. Techniques included Sentiment Analysis, Named Entity Recognition (NER), and demographic profiling of 1,888 active users.
Findings
Descriptive analysis of the collected corpus identifies Saudi Arabia as the primary discourse locus (accounting for approximately 47% of traffic). Semantic analysis reveals a significant discursive shift from economic to cultural themes, termed ‘Entrepreneurial Nationalism,’ highlighted by a substantial growth in references to state initiatives like ‘Vision 2030.’ Within this specific dataset, the ecosystem exhibits ‘fragile positivity’ (an overwhelmingly high positive-to-negative descriptive ratio), indicating a ‘spiral of silence’ regarding critical engagement. Furthermore, user profiling uncovers an ‘elite oligarchy’ dominated by middle-aged technocrats (35–49) and highly educated individuals (e.g., PhD holders), while Generation Z remains largely marginalized.
Research limitations/implications