Search Results (313)

The image of the journalist in popular culture has increasingly added value to metajournalistic discourse. These portrayals have the power to influence the audience’s perception of real-world journalists and the industry. However, most research analyzes portrayals in film and television. Using Terry Pratchett’s fantasy novel “The Truth,” this study explored how journalism, the media industry, and the journalist are portrayed in fantasy literature. Through a textual analysis of the novel, it was found that the work was a celebratory portrayal of journalism that shared a variety of themes found in film and television portrayals. Though its ethics were challenged throughout the novel, the Ankh-Morpork Times was devoted to the truth, served the watchdog role, and practiced social responsibility. Additionally, the novel’s historical rendition of the penny press highlighted the competitiveness of the media industry, how the public interest was challenged by political and corporate influence, and offered a portrayal of naïve news consumers. Lastly, it was found that William de Worde portrayed an ethical journalist and followed the common investigative journalist trope, but his character strayed from the usual editor, publisher, and male reporter tropes found in film and television. This study also suggests the possibility of looking at negative portrayals of journalism in fiction as a series of critical incidents in which journalism has difficulty fully repairing its paradigm. Full article

The proliferation of intelligent sensor networks in urban surveillance and remote sensing has triggered the explosive growth of unstructured visual sensor data. Accurately retrieving targets from these massive streams based on complex cross-modal user intents remains a critical bottleneck for efficient intelligent perception. Composed Image Retrieval (CIR) addresses this by enabling retrieval via a multi-modal query that combines a reference image with semantic control signals. However, existing methods often struggle with abstract instructions in real-world scenarios. Consequently, models often suffer from feature distribution shifts due to focus ambiguity, as well as semantic erosion caused by highly entangled visual and textual features. To address these challenges, we propose a geometry-based Selective Orthogonal Projection Network (SOP). First, the Selective Focus Recovery module quantifies instruction uncertainty via information entropy and calibrates shifted query features to the true target distribution using structural consistency regularization. Second, to ensure data fidelity, we introduce Orthogonal Subspace Projectionand Geometric Composition Fidelity. These mechanisms employ Gram–Schmidt orthogonalization to decouple features into a constant visual base and an orthogonal modification increment, restricting semantic modifications to the null space. Extensive experiments on FashionIQ, Shoes, and CIRR datasets demonstrate that SOP significantly outperforms SOTA methods, offering a novel solution for efficient large-scale sensor data retrieval and analysis. Full article

Text-based person retrieval (TBPR) aims to search for target person images from large-scale video clips or image databases based on textual descriptions. The quality of benchmarks is critical to accurately evaluating TBPR models for their ability in relation to cross-modal matching. However, we find that existing TBPR benchmarks have a common problem, which often leads to ambiguities where multiple images of persons with different identities have very similar or even identical textual descriptions. As a consequence, although TBPR models correctly retrieve the images corresponding to a given description, such matches may be erroneously evaluated as mismatches due to the above annotation problem. We argue that the main cause of this problem is that each person image is annotated individually without reference to other similar images, making it challenging to provide distinctive descriptions for each image. To address this problem, we propose an effective and efficient annotation refinement framework to improve the annotation quality of TBPR benchmarks and thereby mitigate annotation-induced mismatches. Firstly, sets of images prone to mismatches are automatically identified by TBPR models. Then, by leveraging multimodal large language models (MLLMs), multiple images are simultaneously processed and distinctive descriptions are generated for each image. Finally, the original descriptions are replaced to improve the annotation quality. Extensive experiments on three popular TBPR benchmarks (CUHK-PEDES, RSTPReid and ICFG-PEDES) validate the effectiveness of our proposed method for improving the quality of annotations, and demonstrate that the resulting more discriminative captions can truly benefit the mainstream TBPR models. The improved annotations of these benchmarks will be released publicly. Full article

This article re-examines Gen 28:10–22 through a diachronic analysis informed by its close literary and thematic parallels with Gen 35:9–15. In light of recent developments in Pentateuchal scholarship that question the traditional dating of supposedly pre-Priestly texts, the study adopts a method grounded in detailed textual, linguistic, and literary observation rather than reliance on fixed source-critical models. The analysis argues that Gen 28:10–22 is not a unified narrative but a composite text consisting of an early narrative core overlaid by a post-Priestly addition. Particular attention is given to the ritual acts of pillar erection, anointing with oil, and Jacob’s vow, which exhibit strong affinities with Priestly and Deuteronomistic idioms. A comparison with the Priestly account in Gen 35 suggests that the post-Priestly expansion in Gen 28 responds polemically to a Priestly tendency to neutralize Bethel’s sanctity. The article situates this literary development within the religious landscape of Persian-period Yehud. Full article

In the context of polarized media discourse, this study examines how outlets with distinct political leanings constructed multimodal representations of the 2025 Los Angeles protests. Adopting a corpus-assisted Multimodal Critical Discourse Analysis (MCDA) framework, this study integrates Entman’s framing theory with Kress and van Leeuwen’s visual grammar to analyze news coverage of the protests. The results reveal a divergence in multimodal strategies. Fox News employs a delegitimization frame through a dominant strategy of reinforcement, where images serve as direct evidence for textual claims. Conversely, CNN constructs a conditional legitimacy frame via a more nuanced strategy, through which the outlet strategically utilizes multimodal contradiction to negotiate with the “protest paradigm” and mitigate the visual reality of disorder. The findings demonstrate how partisan media leverage distinct multimodal strategies to reconstruct opposing social realities. The study contributes to political discourse research by going beyond textual bias to reveal how multimodal strategies function in media polarization environments. Full article

Large Language Models (LLMs) have demonstrated extraordinary capabilities in natural language generation; however, their real-world deployment is frequently hindered by the generation of factually incorrect or biased content, along with an inherent deficiency in transparency. To address these critical limitations and thereby enhance the reliability and explainability of LLM outputs, this study proposes a novel integrated framework, namely the Adaptive Knowledge-Driven Correction Network (AKDC-Net), which incorporates three core algorithmic innovations. Firstly, the Hierarchical Uncertainty-Aware Bias Detector (HUABD) performs multi-level linguistic analysis (lexical, syntactic, semantic, and pragmatic) and, for the first time, decomposes predictive uncertainty into epistemic and aleatoric components. This decomposition enables principled, interpretable bias detection with clear theoretical underpinnings. Secondly, the Neural-Symbolic Knowledge Graph Enhanced Corrector (NSKGEC) integrates a temporal graph neural network with a differentiable symbolic reasoning module, facilitating logically consistent and factually grounded corrections based on dynamically updated knowledge sources. Thirdly, the Contrastive Learning-driven Multimodal Explanation Generator (CLMEG) leverages a cross-modal attention mechanism within a contrastive learning paradigm to generate coherent, high-quality textual and visual explanations that enhance the interpretability of LLM outputs. Extensive evaluations were conducted on a challenging medical domain dataset to validate the effectiveness of the proposed AKDC-Net framework. Experimental results demonstrate significant improvements over state-of-the-art baselines: specifically, a 14.1% increase in the F1-score for bias detection, a 19.4% enhancement in correction quality, and a 31.4% rise in user trust scores. These findings establish a new benchmark for the development of more trustworthy and transparent artificial intelligence (AI) systems, laying a solid foundation for the broader and more reliable application of LLMs in high-stakes domains. Full article

While scholarship has predominantly focused on the “harmony” of Confucian ethics or the functional and generative “harmony” of pre-Qin Lao-Zhuang Daoism, this study identifies a unique conceptual system of “harmony” in Han Dynasty Daoism through a textual excavation of Yan Zun’s Laozi zhigui. Yan Zun transcends the relatively abstract generative narratives of pre-Qin Daoism by creatively substantializing “harmony” into “supreme harmony”, positioning it as a pivotal stage in the four-tiered cosmogonic schema of “Dao–De–Shenming–supreme harmony”. By regarding “supreme harmony” as the “ancestor of Heaven and Earth” and the ontological foundation for the nature and life of all things, Yan Zun endows “harmony” with a definitive ontological status. This cosmological and ontological category further permeates the domains of self-cultivation and state governance. In the realm of self-cultivation, Yan Zun advocates for “valuing the body and nourishing the spirit”, promoting the practice of spirit and qi embracing and tranquil non-action to achieve the existential realization and transcendence of individual life; in the realm of state governance, he criticizes rites and laws for harming natural harmony, proposing that the ruler should “embody the Dao and tread upon harmony”. This approach establishes a governance of non-action that aligns with the “utmost softness” of supreme harmony, thereby reconstructing an ideal political order where “harmonious qi flows freely.” The concept of “supreme harmony” advocated by Yan Zun not only marks the maturation of Han Daoist qi-cosmology, but also offers a new theoretical horizon for re-understanding the transformation of the concept of “harmony” from ethics to ontology in Chinese philosophy. Full article

Financial document understanding remains a critical challenge for Large Language Models, primarily due to the complex interplay between narrative text and structured numerical tables. Existing Retrieval-Augmented Generation (RAG) systems often treat these modalities in isolation, leading to significant failures in tasks requiring joint reasoning. This study introduces HierFinRAG, a novel hierarchical multimodal framework designed to unify tabular and textual data processing. Our approach employs a Table-Text Graph Neural Network (TTGNN) to explicitly model semantic and structural dependencies between table cells and corresponding text, coupled with a Symbolic–Neural Fusion module that routes queries between a neural generator and a symbolic calculator for precise arithmetic operations. We evaluate the system on the FinQA and FinanceBench datasets, comparing performance against strong baselines including Vanilla RAG and GPT-4o with Code Interpreter. Results demonstrate that HierFinRAG achieves an Exact Match score of 82.5% on FinQA, surpassing the best baseline by 6.5 percentage points, while maintaining a 3.5× faster inference latency than agentic approaches. These findings indicate that integrating hierarchical structural awareness with hybrid reasoning significantly enhances the accuracy and interpretability of financial artificial intelligence systems. Full article

In disaster emergency response, spatial location information embedded within social media texts holds substantial value for the rapid localization of affected areas and the implementation of precise rescue operations. Existing research predominantly employs natural language processing and deep learning technologies for geographic information extraction; however, two critical limitations persist: first, insufficient integration of textual semantic features for disaster relevance determination, resulting in inadequate correlation between extracted results and actual disaster locations; second, absence of mechanisms for identifying affected sites in multi-location contexts, thereby compromising decision support efficacy. Addressing these challenges, this study proposes a hierarchical disaster location information extraction framework that integrates semantic understanding. The framework operates through a three-tier hierarchy: data-level adversarial augmentation, semantic-level dynamic parsing, and parameter-level scale optimization. It achieves three core functionalities: (1) precise determination of disaster relevance for geographic location information; (2) identification of affected areas in multi-location contexts; (3) establishment of a logarithmic scaling relationship between LLM parameter scale and optimal prompt sample size. Full article

Airplanes are the most popular investigation objects as a dynamic and critical component in remote sensing images. Accurately identifying and monitoring airplane behaviors is crucial for effective air traffic management. However, existing methods for interpreting fine-grained airplanes in remote sensing data depend heavily on large annotated datasets, which are both time-consuming and prone to errors due to the detailed nature of labeling individual points. In this paper, we introduce Text2AIRS, a novel method that generates fine-grained and realistic Airplane Images in Remote Sensing from textual descriptions. Text2AIRS significantly simplifies the process of generating diverse aircraft types, requiring limited texts and allowing for extensive variability in the generated images. Specifically, Text2AIRS is the first to incorporate ground sample distance into the text-to-image stable diffusion model, both at the data and feature levels. Extensive experiments demonstrate our Text2AIRS surpasses the state-of-the-art by a large margin on the Fair1M benchmark dataset. Furthermore, utilizing the fine-grained airplane images generated by Text2AIRS, the existing SOTA object detector achieves 6.12% performance improvement, showing the practical impact of our approach. Full article

Nahj al-Balāghah is widely recognized as a foundational and authoritative scripture in Shia Islam. One notable aspect of Nahj al-Balāghah is the deliberate selection and structured arrangement of Ḥadiths. According to the book’s introduction, al-Raḍī explains that he chose the Ḥadiths based on literary considerations. An analysis comparing the selected Ḥadiths with their full versions suggests their inclusion was determined not only by eloquence and rhetorical value but also by conceptual significance. Through textual and descriptive analytical methods, this study examines the author’s motives, especially his political and religious aims, in incorporating materials related to ʿĀisha. A comparison of the relevant ḥadīths in Nahj al-Balāghah and other historical sources indicates that Sayyid Raḍī omitted—or at least refrained from including—certain statements attributed to ʿAlī regarding the Prophet Muḥammad’s youngest wife. The omitted parts concern ʿĀisha’s inconsiderate behavior, grudges, sins, following Satan, and ignoring the Prophet’s prediction. Considering sectarian conflicts between Shiites and Sunnis in the 3rd and 4th centuries AH, some arising from criticisms of ʿĀisha’s conduct and sometimes escalating into violence, al-Raḍī, the supreme judge appointed by the ʿAbbāsid Caliphate, was compelled to omit and censor ʿAli’s harsh remarks about ʿĀisha to prevent further sectarian tensions. Full article

The accurate and reliable evaluation of large language models (LLMs) in medical domains is critical for real-world clinical deployment, automated medical reasoning, and patient safety. However, the evaluation process is highly sensitive to prompt design, and prevalent reliance on fixed or randomly sampled prompt policies often fails to dynamically adapt to clinical context, question complexity, or evolving safety requirements. This article presents a novel reinforcement learning-based framework for multi-prompt selection, which dynamically optimizes prompt policy per input for medical LLM evaluation across the Medical Knowledge Question-Answering dataset (MKQA), the Medical Multiple-Choice Question dataset (MCQ), and the Doctor-Patient Dialogue dataset. We formulate prompt selection as a Markov Decision Process (MDP) and employ a deep Q-Network (DQN) agent to maximize a reward signal incorporating textual accuracy, domain terminology coverage, safety, and dialogue relevance. Experiments on three medical LLM benchmarks demonstrate consistent improvements in composite reward (e.g., a 6.66% increase in MKQA vs. Random Baseline, and a 2.41% increase in Dialogue vs. Fixed Baseline) when compared to baselines. This was accompanied by robust enhancements in Safety (e.g., achieving 1.0000 in MKQA, a 5.26% increase vs. Fixed Baseline; and a 5.03% increase in Dialogue vs. Fixed Baseline) and substantial gains in Medical Terminology Coverage (e.g., a 74.61% increase in MKQA vs. Fixed Baseline, and a 9.13% increase in MCQ vs. Fixed Baseline) when compared to baselines. While varying across tasks, an improvement in accuracy was observed in the MKQA task, and the framework effectively optimizes the multi-objective reward function, even when minor trade-offs in other metrics like Accuracy and Contextual Relevance were observed in some contexts. Our framework enables robust, context-aware, and adaptive evaluation, laying a foundation for safer and more reliable LLM application in healthcare. Full article

In the era of globalization, Chinese firms increasingly leverage sequential cross-border mergers and acquisitions to navigate complex international environments. Understanding the sustained impact of this strategy requires a holistic view beyond isolated transactions. Adopting an open systems perspective, this study examines how sequential cross-border M&As influence the evolution of firms’ innovation boundaries as a dynamic system property. Combining grounded theory with textual analysis of Chinese M&A cases, this study develops an integrated systemic process framework and empirically tests it using data from Chinese listed firms (2002–2021) via fixed-effects models. Results reveal that sequential cross-border M&As act as external inputs that trigger internal system reconfiguration, significantly expanding innovation boundaries. This expansion process is mediated by dynamic capabilities, which constitute the firm’s core adaptive mechanism, and is moderated by resource slack that functions as critical system redundancy. These findings contribute to systems science by elucidating how firms, conceptualized as complex adaptive systems, transition from isolated deal-making to sustained capability building through iterative feedback loops in global competition. Full article

This article reinterprets modern realist drama as a site of secular spirituality, where aesthetic form sustains the sacred under conditions of modern secularity. Employing a phenomenological–theological framework, it integrates Charles Taylor’s account of the secular age, Mircea Eliade’s sacred–profane dialectic and hierophany, and René Girard’s anthropology of sacrifice. Through textual and performance-historical analysis of key works—Ibsen’s A Doll’s House (1879) and An Enemy of the People (1882)—together with Chinese modern drama shaped by Ibsenization, including Hu Shi’s translations, Lu Xun’s critiques, and Cao Yu’s Thunderstorm (1934), the article argues that realist theatre fulfils religious functions in secular culture: revelation as truth-telling, confession as critical self-disclosure, and renewal as ethical transformation. In early twentieth-century China, the encounter between Ibsen’s moral realism and indigenous moral traditions generated a distinctive spiritual humanism, in which theatre assumed ritual and didactic functions traditionally associated with religious practices. Full article

The farmland protection compensation system plays a pivotal role in addressing the dual global crises of food insecurity and ecological degradation, as well as in overcoming persistent challenges in China’s agricultural governance. By internalizing the opportunity costs borne by stakeholders fulfilling statutory obligations for farmland protection, this mechanism offers effective incentives for their active engagement, thereby establishing a societal-level interest-balancing framework conducive to sustainable land management. Existing research in China has mainly concentrated on empirical analyses of implementation models, regional disparities, and policy effectiveness evaluations of farmland protection compensation schemes. Nevertheless, systematic exploration of the normative construction and improvement pathways of the compensation rules themselves remains relatively underdeveloped. Based on the practical requirements and institutional constraints of China’s current farmland protection compensation regime, this study adopts an integrated approach that combines comparative legal analysis, textual review of regulatory documents, and empirical research to critically examine feasible paths for institutional improvement. The research findings emphasize that the optimization of China’s farmland protection compensation rules should be guided by three core principles: market orientation, ecological sustainability, and precision-based targeting. Specifically, the establishment of scientifically sound methods for calculating compensation amounts is crucial for reconciling the interests of conservation actors with inter-regional development disparities. Meanwhile, the compensation mechanism should be strategically utilized to strengthen positive incentives for ecosystem conservation. Ultimately, such institutional improvement aims to ensure the sustainable utilization of farmland resources while safeguarding global food security and maintaining the Earth’s ecological balance. Full article