assess content credibility

This study addresses the challenge of assessing the credibility of emerging web domains that lack historical reputation, rendering traditional evaluation methods ineffective. To overcome this limitation, the authors propose the Domain Credibility Evaluation Framework (DCEF), which, for the first time, leverages temporal sequences of article-level credibility signals within a domain to emulate the judgment logic of professional fact-checkers and automatically predict the overall trustworthiness of previously unseen domains. Built upon expert-annotated data, DCEF employs time-series modeling and aggregates article-level credibility scores into an end-to-end automated assessment system. Experimental results demonstrate that domain-level credibility can be effectively predicted using only article content, establishing a novel paradigm and a practical pathway for evaluating the trustworthiness of emerging domains.

This work addresses the prevalence of misleading content in social media feeds, which undermines overall information credibility. The authors propose a Pareto-optimal reranking framework that simultaneously maximizes credibility enhancement while preserving fidelity to the original ranking. The approach integrates a bi-objective optimization model based on Spearman footrule distance with a semi-automated credibility scoring mechanism, which combines human fact-checking, community notes, and retrieval-augmented generation techniques. Experiments on real-world data from the X platform demonstrate that the reranked results deviate by no more than 7% from the Pareto front and support flexible adaptation to diverse credibility signals, making the framework readily applicable across different platform requirements.

This study reveals that large language models struggle to effectively assess the veracity of statistical evidence when integrating multi-source information, exhibiting a tendency to rely on superficial stylistic cues in methodological text rather than numerical plausibility when judging source credibility. The work identifies a previously undocumented “cognitive alignment” bias—where models prefer sources with strong analytical register over those with content consistency. Employing interpretable techniques including causal tracing, linear probing (AUC: 0.83–0.92), and component-level attribution, the authors replicate this blind spot across five mainstream models through cross-model and cross-domain experiments. Further analysis localizes the issue to a methodology-register gating mechanism and demonstrates that neither prompt engineering nor post-training interventions adequately mitigate the bias, instead raising concerns about model generalization.

This work addresses the efficiency bottleneck of manual reproducibility reviews in safety-critical domains such as the Internet of Things and cyber-physical systems, which hampers research transparency and deployability. The paper presents the first systematic framework leveraging large language models (LLMs) to automate reproducibility assessment by integrating natural language understanding, code generation, sandboxed environment auto-configuration, and rule-guided flaw detection. This approach enables reproducibility scoring, automatic execution environment setup, and identification of methodological flaws. Experimental results demonstrate that the proposed method achieves over 72% accuracy in reproducibility judgment, automatically constructs executable environments for 28% of runnable artifacts, and attains F1 scores exceeding 92% across seven common categories of methodological defects, substantially enhancing both the efficiency and quality of reproducibility review.

This study investigates how different visual content formats—specifically image-text combinations, infographics, and data visualizations compared to plain text—affect perceived credibility of social media posts. Grounded in processing fluency theory, a preregistered large-scale online experiment (N = 1200) coupled with structural equation modeling reveals that image-text and infographics significantly enhance credibility, whereas data visualizations do not produce such an effect. Aesthetic appeal indirectly boosts credibility by increasing processing fluency, while production quality shows no significant impact. This work is the first to differentiate the unique effects of distinct visual formats on credibility judgments, establishes processing fluency as a key mediating mechanism, and redefines the theoretical role of visual features in multimodal credibility assessment.

This work addresses systematic limitations in existing creative quality alignment (CQA) datasets, particularly their inadequate modeling of audience preferences and insufficient coverage of real-world logical constraints. To overcome these issues under stringent engineering and data scarcity conditions, the authors propose a low-resource CQA approach that leverages only around one hundred expert-annotated chain-of-thought (CoT) examples. By uncovering a dual mechanism between appreciation and generation tasks within conditional generative architectures, the method enables automatic transfer of calibrated knowledge from the appreciation module to the generation module. Experimental results demonstrate that the proposed framework substantially mitigates the shortcomings of current datasets and validates the practical feasibility of aligning generative models with nuanced creative quality metrics in real-world engineering settings.

This study addresses growing concerns among content creators who, misinterpreting binary AI labels as fine-grained tracking tools, often proactively erase digital traces to safeguard privacy and reputation—thereby undermining the traceability and safety mechanisms of AI-generated content (AIGC). Through semi-structured interviews with 21 creators and empirical testing of 16 user-reported image modification techniques across six generative platforms, the research combines qualitative insights with quantitative evaluation of underlying provenance methods, including watermarks and metadata. Findings reveal that defensive de-identification practices—such as coarse quantization—significantly degrade the accuracy of current AI detectors, with substantial inter-platform variability and notable false positives on human-created images. The work thus calls for resilient, implicit AI labeling workflows aligned with creators’ incentive structures.

Scientific peer review often includes subjective or unverifiable claims that compromise its fairness. This work proposes the first end-to-end claim verification system tailored for peer review comments. The system extracts verifiable claims from reviews, retrieves and re-ranks relevant evidence from the manuscript, performs fact-checking via natural language inference, and presents results through an interactive visual interface. Designed with modularity in mind, it allows flexible substitution of retrieval, re-ranking, and inference components, making it adaptable for reviewers, authors, and program committees alike. The system is publicly accessible via a live demo platform and API, accompanied by tutorials, significantly enhancing the transparency and verifiability of the peer review process.

This work addresses the critical limitation of existing open-domain research report generation methods, which often lack effective mechanisms to assess content credibility, thereby risking hallucination and misinformation. To mitigate this, the paper introduces a Deep Research Agent that, for the first time in settings without ground-truth references, incorporates a progressive confidence estimation and calibration framework. By leveraging deep retrieval and multi-hop reasoning, the agent anchors each claim to verifiable evidence and assigns an interpretable confidence score. Integrating a structured workflow with cognitive modeling of confidence, the proposed approach substantially enhances the transparency, interpretability, and user trustworthiness of generated reports while effectively suppressing hallucinatory and misleading content.