Sage University

The increasing availability of multilingual and multimodal data on online platforms has created a need for the development of sophisticated models for cross-lingual sentiment analysis. Although existing sentiment analysis models are largely dependent on monolingual text data, they tend to fail in generalizing across languages and modalities, especially in low-resource and code-switching settings. This work aims to handle the most important task of combining heterogeneous modalities of text and audio data in multiple languages for accurate and interpretable sentiment classification. X-Sent-Fuse, a cross-lingual sentiment fusion model using Transformers, combines text and audio modalities to make accurate, robust, and culturally informed sentiment predictions for a wide range of languages. The architecture incorporates a Multimodal Transformer Cross-Attentive Gated Fusion (MT-CAG2F) module to effectively align semantic and prosodic features, followed by a BiSRO-Net classifier, to ensure stable and high-performing predictions. A lightweight ensemble of multilingual large language models further improves linguistic adaptability. Extensive evaluations on four benchmark datasets, Sentiment analysis in Twitter, Amazon Reviews, mTEDx, and CMU-MOSEAS, demonstrate state-of-the-art performance, achieving up to 98.67

Doxorubicin, a secondary metabolite of Streptomyces peucetius var. caesius and a member of the anthracycline family, exerts anticancer effects via DNA intercalation and topoisomerase II inhibition in tumor cells. However, its clinical application is limited by dose-dependent and cumulative cardiotoxicity. The mechanisms underlying doxorubicin-induced cardiotoxicity (DIC) include oxidative stress, lipid peroxidation, mitochondrial dysfunction, calcium dysregulation, disrupted iron homeostasis, nitric oxide release, and inflammatory mediator production. Emerging evidence highlights autophagy dysregulation, with doxorubicin upregulating cardiac autophagy by suppressing GATA4 and ribosomal protein S6 kinase beta-1(S6K1). Mitochondria-dependent ferroptosis also plays a significant role, driven by downregulation of glutathione peroxidase 4 (GPX4), lipid peroxidation via DOX-Fe2+ complexes, and dysregulated iron metabolism. Additionally, DOX triggers pyroptosis in cardiomyocytes, involving proteins such as NLRP3 (NOD-, LRR-, and pyrin domain-containing protein 3), caspase-3, and gasdermin D (GSDMD). Epigenetic alterations, including DNA hypomethylation (via downregulation of DNMT1 (DNA (cytosine-5)-methyltransferase 1), changes in microRNA levels (e.g., upregulation of miR-520h targeting HDAC19 (histone deacetylase 1), and histone deacetylase inhibition, exacerbate cardiac damage. Recent studies also emphasize the role of gut microbiota in doxorubicin-induced cardiotoxicity. Doxorubicin induces dysbiosis, leading to cardiomyocyte apoptosis and elevated myocardial enzyme levels. Interventions such as dietary modifications, fecal microbiota transplantation, probiotics, and natural compounds like glabridin and emodin show promise. Glabridin reduces inflammation by modulating colonic macrophage polarization, while emodin inhibits ferroptosis via gut microbiota remodeling mediated by Nrf2. This review explores oxidative stress, lipid peroxidation, ferroptosis, apoptosis, inflammation, autophagy, epigenetics, and gut microbiota in DIC, alongside promising pharmacological strategies to mitigate its effects.

Optical control of ferroic order offers a transformative route for manipulating magnetic states in multiferroic systems, especially at ferromagnetic/ferroelectric (FM/FE) interfaces. However, achieving stable, reversible, and room-temperature modulation of magnetism via light remains a key challenge, primarily due to limited coupling efficiency and poor structural control at the nanoscale. This limitation can be overcome by utilizing strain-mediated coupling mechanisms that convert photostrictive responses in the FE layer into magnetic reorientation in the FM layer. Here, a reversible, light-induced strain effect is demonstrated in PMN-PT single crystals, where domain variants visualized via Piezoresponse Force Microscopy undergo controlled reconfiguration under visible laser illumination. X-ray diffraction reciprocal space maps confirm lattice deformation, while Raman spectroscopy reveals local structural changes. In Fe/PMN-PT heterostructures, this optically induced strain drives magnetic axis reorientation in the Fe layer, confirming robust magneto-electric coupling. The discovery establishes a non-contact, energy-efficient approach to manipulate magnetism, advancing the field toward ultrafast, reconfigurable magneto-optical devices.

PurposeOrganizations often fail to implement strategies internally when they do not effectively orchestrate their resources and capability. Therefore, this study develops a scale for measuring enterprise architecture (EA) capability and establishes a framework for evaluating its components. Properly assessing EA capability enables alignment between business and information technology (IT).Design/methodology/approachA comprehensive literature review was conducted along with expert interviews to determine which systems comprise EA capability. A questionnaire was developed and distributed to 131 firms and convergent and discriminant validity analyses based on partial least squares structural equation modeling were conducted. The proposed model includes 17 characteristics associated with the strategic attributes, the business model that forms the business construct, and the system and technology models that inform the IT construct.FindingsThe results demonstrate that EA capability can orchestrate business and IT, and that IT resources are crucial to building EA capability. Additionally, it was revealed that when aligned with the firm's objectives and business model EA capability can leverage organizational agility, enabling changes through innovation or operational adjustments.Originality/valueThis study addresses the lack of methods for evaluating EA capability within organizations by providing empirical evidence to assess their effectiveness in orchestrating business and IT resources. This study complements the results of previous research by considering IT's influence on resource orchestration theory to enable business.

Data-driven crop disease diagnosis has become an option since the recent accelerated digitization of agriculture and various IoT-enabled devices and UAVs were implemented. Nonetheless, the problem in the centralised deep learning models of the area related to high data ownership, privacy, and transmission overheads is quite real, particularly in geographically distributed farming belts. To overcome these, this study suggests the use of Federated Agricultural Disease Diagnosis Network (FedAgri-Net), a privacy preserving crop disease prediction system on distributed agricultural zones by utilization of federated learning. Local nodes of farms in this architecture train disease classifiers using lightweight convolutional neural networks, on local images, without exchanging raw inputs. Gradient updates are in an encrypted form and sent to nearest edge aggregators that perform the coordination of model synchronization and passes processed updates to a central cloud server that performs global aggregation. To improve accuracy in the heterogeneous areas, the framework enables individual model adjustment and identification of anomalies to manage non-IID data distributions. Moreover, built-in explanatory processes give visual explanations of disease symptoms and can be used by farmers as a guide to trust and decision-making. Practical simulations prove that FedAgri-Net is a protocol with a high performance of diagnostic and data independence. The suggested FedAgri-Net approach has an overall accuracy of 94.5 %.