In the rapidly evolving digital landscape, the proliferation of online image and video sharing has reached unprecedented levels. This surge has been accompanied by the emergence of deepfake technology, powered by generative adversarial networks and deep learning techniques, enabling the creation of highly realistic fabricated images and videos. Consequently, the widespread dissemination of deepfake content on social media platforms has prompted the global community to seek effective methods for detecting and combating this issue. Recognizing that Artificial Intelligence (AI) has played a significant role in propagating deepfake challenges, researchers and experts believe that AI can also provide viable solutions. In this study, we explore the application of AI for deepfake image detection. Specifically, we focus on three convolutional neural network (CNN) algorithms—VGG16, VGG19, and ResNet—for this purpose. To evaluate the performance of these CNN models, a dataset comprising 1,200 images, including a combination of fake and real images, was utilized. The deepfake images were generated using FaceApp, a prominent tool for creating manipulated visuals. Our findings demonstrate that VGG19 outperforms both VGG16 and ResNet50, achieving an impressive accuracy rate of 98% on the test dataset when small size is small. By harnessing the power of AI, particularly through the application of CNN algorithms, this research makes a significant contribution to the field of deepfake image detection. The results underscore the efficacy of VGG19 in accurately identifying manipulated images, thereby providing valuable insights for the development of robust detection systems to combat the proliferation of deepfakes in today's digital world.

Recently, there has been a significant increase in the number of Artificial Intelligence (AI) systems, software, and components. As a result, it is crucial to evaluate their quality. Quality models for AI have in fact been proposed, but there is a lack of Systematic Mapping Studies (SMS) for quality models in AI systems, software, and components. The goal of this paper is to understand, classify, and critically evaluate existing quality models for AI systems, software, and components. This study conducts an SMS to investigate quality models proposed by various authors in the past. The study only found quality models for AI systems and software. So far, the SMS has revealed no work on AI software component quality models. Finally, the limitations of the quality models and the implications for future research and development efforts are discussed.

Objective: The objective of this study is to conduct a comprehensive bibliometric analysis of greenwashing research from 1995 to 2024, with the aim of understanding the evolution, key themes, and influential contributors in the field. Theoretical Framework: This study uses social network theory and Lotka’s Law to identify key contributors and collaboration patterns in greenwashing research, highlighting influential authors and the concentration of productivity in the field. Method: Scopus was selected as the database for data collection, and data was gathered through systematic searches. The method employed is bibliometric analysis, focusing on performance analysis and science mapping with metrics such as total citations, citations per year, and author dominance. The analysis also explored trends, key contributors, collaboration networks, and etc. Results: The results reveal significant growth in greenwashing research since 2010, with the USA and China identified as major research hubs. Two primary research clusters were identified: corporate sustainability and consumer marketing. The analysis also identified key authors, influential journals, the most impactful articles, collaboration patterns, and changes in research focus over time. Research Implications: The practical and theoretical implications of this research include insights into how greenwashing practices impact corporate reputation, consumer trust, and regulatory frameworks. These implications are relevant for policymakers, researchers, and practitioners aiming to foster transparency and accountability in sustainability efforts. Originality/Value: This study contributes to the literature by providing an in-depth bibliometric analysis that identifies influential contributors, emerging trends, and collaboration patterns in greenwashing research. The relevance of this study is demonstrated by its potential to guide future research and inform policy on addressing deceptive sustainability practices.

Objective: This study aims to investigate the signaling effects of Sustainability Reporting (SR) in listed Singapore firms, particularly in demonstrating firm capitalization and sustainability performance. The research focuses on identifying distinct SR signals using advanced text mining and classification techniques, and their alignment with the United Nations' Sustainable Development Goals (SDGs). Theoretical Framework: The study is grounded in signaling theory, which posits that firms communicate observable characteristics (SR content) to signal unobservable attributes (firm capabilities and sustainability performance). By applying this framework, we analyze how SR may act as a signaling mechanism for large-cap firms and sustainability leaders. Method: This research collects 1,844 sustainability reports from listed Singapore firms between 2016 and 2022. Using a combination of Structural Topic Modeling (STM) and Random Forest (RF) classification models, we extracted and analyzed the key SR signals that differentiate firms by size and sustainability performance. Results: The results indicate that SR signals such as community engagement, climate risk, and compliance are effective in distinguishing between large firms and firms with strong sustainability performance. The Random Forest models achieved high accuracy in identifying these groups. Community engagement, compliance with the Global Reporting Initiative (GRI), and sustainable construction were among the most impactful topics, with large firms focusing more on scope and governance, while sustainability leaders emphasized environmental risks and compliance. Research Implications: The findings provide insights into the practical and theoretical implications of SR as a signaling tool. Key SR topics are identified as critical signals for firms with large capitalization and high sustainability performance, guiding firms in utilizing SR in the context of SDGs. Originality/Value: The findings show that SR topics such as community engagement, climate risk, and compliance serve as key signals for differentiating firms by size and sustainability performance. Firms can strategically emphasize these areas to enhance their SR practices while contributing to broader SDGs.The findings provide insights into the practical and theoretical implications of SR as a signaling tool. Key SR topics are identified as critical signals for firms with large capitalization and high sustainability performance, guiding firms in enhancing their SR strategies.

Ensemble learning is a widely recognized technique in Artificial Intelligence that boosts model performance by combining predictions from multiple classifiers. While traditional ensemble methods effectively combine classifiers within the same domain, they face challenges when integrating models that handle different tasks. This study introduces Conditional Max-Preserving Normalization, a novel approach that extends ensemble methods' applicability across diverse classification domains. Unlike altering deep learning architectures, this method focuses on preserving the most significant prediction while proportionally scaling others to ensure consistency in the combined output. The study utilized the SoftMax function to emulate classification tasks, generating probability vectors for both Human-Car and Cat-Dog classifications. The proposed method identifies the highest confidence value in the combined vector, counts its occurrences, sums the remaining values, and computes a Scale Rate to normalize the vector. The competitive evaluation demonstrated that Conditional Max-Preserving Normalization outperforms traditional ensemble methods in maintaining accuracy and reliability across diverse classification tasks. Formal verification using the Z3 solver affirmed the method's robustness, confirming that the combined vector maintains a valid probability distribution and retains the maximum value. Future research could focus on refining the method to eliminate conditions during normalization, adapting it for binary classification, exploring its application in sequential classification tasks, and extending its use to regression problems. This research lays the groundwork for more robust and adaptable ensemble learning models with potential applications in various real-world scenarios.

Requirements engineering is crucial for software project success. Issues like requirements ambiguity, inconsistency, and unverifiability contribute to unclear, conflicting, or untestable specifications, which can undermine the effectiveness and success of a software project. These issues have been identified as factors contributing to software project failure. However, there’s limited research on the current state of these issues in China. The research objectives of this study are to address the most commonly used methods for expressing Chinese software requirements and uncover issues related to ambiguity, inconsistency, and unverifiability, which can be solved by using artificial intelligence techniques to investigate possible solutions to these problems. An online survey of 422 software professionals in China identifies key issues in Chinese software requirement expressions that AI techniques can address. The study examines various expression methods, tools for enhancing clarity, and challenges specific to Chinese requirements. Findings reveal that ambiguity, inconsistency, and unverifiability significantly impact project success. While natural language specification and prototyping improve clarity, they may increase the time required for requirements engineering. Effective communication is typically achieved through natural language, prototyping, storyboarding, and pseudo-coding, whereas decision tables and block diagrams are less commonly used and linked to problematic requirements. Using tables, prototype diagrams, and natural language descriptions helps mitigate these issues, though it may extend engineering time. The study suggests strategies to improve the efficiency and quality of requirements expression and highlights the need to develop Chinese boilerplates and refining tools to enhance clarity in the future.