Arab Open University

Arabic aspect-based sentiment analysis (ABSA) remains underexplored, particularly end-to-end dual-task learning that jointly performs aspect term extraction (ATE) and aspect sentiment classification (ASC). Most existing Arabic ABSA approaches are pipeline-based or restricted to single-task identification, which weakens task synergy, constrains performance, and rarely leverages explicit syntactic structure. To address this gap, we propose a structure-aware, end-to-end model that integrates contextual semantics with linguistic structure to perform ATE and ASC tasks simultaneously. Our architecture builds on contextual representations from an Arabic BERT, augmented with token-level attention and sequential modeling via a BiLSTM. A gated fusion module then learns adaptive weights to combine these features into a unified representation. To incorporate linguistic relations, we introduce a dual-branch graph convolutional network that captures (i) dependency-based syntactic relations and (ii) part-of-speech relational patterns. Both branches use attention-weighted adjacency matrices to reduce sensitivity to parser noise and to emphasize sentiment-relevant links. The proposed model is evaluated on three Arabic ABSA datasets (Hotel, News, and HAAD). It achieves ATE and ASC accuracies of 95.29% and 90.87% on Hotel, 91.85% and 82.15% on News, and 90.96% and 75.06% on HAAD, respectively. Under strict end-to-end exact-match evaluation of aspect-sentiment pairs, the model achieves F1-scores of 71.12%, 64.98%, and 33.81% on the Hotel, News, and HAAD datasets, respectively.

Ultrafiltration (UF) is widely used in the biorefinery process but is not well represented in common simulation tools. Due to the lack of a broadly applicable model, a model was developed and simulated in ProsimPlus®. The model has been used to describe the performance of an ultrafiltration membrane for concentrating phenolic compounds under polarization effects. This article demonstrates how the integration of process system engineering and life cycle assessment provides an assessment of environmental impacts without relying on time-consuming and costly experiments. Therefore, this paper proposes an eco-design tool for the biorefinery process by coupling ultrafiltration modeling with environmental impact assessment. The parameters were identified for subsequent simulation in Simapro® Developer to quantify the environmental impact of the ultrafiltration process. The ultrafiltration system involves three key stages: cleaning, filtration, and infrastructure. The results show that filtration is the main contributor to the environmental impact, but cleaning frequency is important in determining the relative impact of each stage. The study also demonstrates that optimizing operating parameters, particularly pressure and tangential velocity, reduces the need for membrane modules, thereby reducing the impact of cleaning, filtration, and infrastructure. The model was evaluated under different operating conditions, with Reynolds numbers ranging from 4000 to 6000 and pressures from 0.8 to 2 bar. The lowest environmental impact was obtained at 2 bar and Re = 5500. Under these optimized conditions, the environmental impacts of the filtration and cleaning stages were reduced by up to 55% and 72%, respectively. These results demonstrate that determining the number of modules and the cleaning frequency are crucial for the proper application of this approach.

This study examines the relationship between sustainability performance, eco-friendly production, and board characteristics and their impact on the green revenues of industrial firms. Using a panel dataset of 6084 firm-year observations from multiple countries (2018-2023), the study employs panel data analysis to assess these relationships. A dummy variable approach differentiates firms with green revenues from those without, followed by a segmented analysis of high- and low-performing green firms. The results indicate that sustainability performance positively influences green revenues, while board size and diversity have negative effects. Conversely, board expertise and sustainability compensations enhance green revenues. Interestingly, the findings highlight that high-performing green firms adopt proactive sustainability strategies, whereas low-performing firms rely on incentives. The study contributes to the literature by providing evidence from industrial firms across developed and developing countries, offering policy implications for sustainability governance and corporate incentives.

Purpose This study aims to investigate the effect of digital transformation on accounting conservatism, emphasizing the moderating role of industry type in an emerging market context. Focusing on Egypt, it examines whether differences in sectoral digital maturity shape how technological adoption influences conservative financial reporting practices, thereby highlighting the importance of institutional and operational heterogeneity across industries. Design/methodology/approach Drawing on signaling theory, agency theory and the resource-based view, the study employs a quantitative research design using panel data from non-financial firms listed on the Egyptian Exchange (EGX) over the period 2019–2023. The empirical analysis relies on multivariate regression models to test the direct effect of digital transformation and its interaction with industry characteristics. To address potential endogeneity concerns, the study applies two-step System GMM estimations, ensuring robust and reliable inferences. Findings The results indicate that digital transformation is positively associated with accounting conservatism. This relationship is significantly stronger in industries characterized by higher levels of digital maturity, suggesting that sectoral readiness conditions the extent to which firms can translate digital investments into improved governance outcomes. Overall, industry digital maturity amplifies the conservatism-enhancing effect of digital transformation. Practical implications The findings suggest that managers can strengthen conservative reporting and internal control systems through targeted digital investments. Regulators and auditors should account for industry-specific digital maturity when evaluating reporting quality and disclosure risk. Firms in technology-intensive sectors may gain particular benefits from prioritizing advanced digital capabilities aligned with governance objectives. Social implications By enhancing transparency and reducing information asymmetry, digital transformation supports accountability and investor confidence, which is especially critical in emerging markets with developing institutional frameworks. Originality/value This study contributes to the accounting literature by introducing industry type as a moderating factor in the digital transformation–conservatism nexus and by employing advanced GMM techniques. It offers context-sensitive evidence from Egypt and extends conservatism research through the use of digitalization measures derived from content analysis.

This study examines students’ perceptions of the rebranding initiative at Arab Open University (AOU), Kuwait Branch, with a particular focus on the newly introduced logo and visual identity. Using a cross-sectional quantitative survey design, data were collected from undergraduate students across different faculties. The findings indicate generally positive evaluations of the new logo in terms of modernity, professionalism, visual appeal, and alignment with institutional values. Inferential analysis further revealed a strong positive association between overall logo evaluation and students’ recommendation intention (r = .80, p < .001), with regression results showing that logo perception explained approximately 65