Generation Of Tools Research Articles

Students’ perception of the impact of AI generative tools in learning the English language

The advancement of AI techniques has presented several challenges for teachers and learners in language classrooms. Yet, students have demonstrated rapid improvements when given direct opportunities to experiment with advanced software. Despite AI’s substantial progress, the core of teacher-learner exchanges remains deeply anchored in human intelligence. As technology becomes more integrated into language classrooms, AI needs to imitate and complement human interactions effectively. This study aimed to investigate students' perceptions of generative tools’ impacts on their writing abilities. A quantitative design was employed, and the research was conducted at the National University of Cheasim Kamchaymear. The study involved 120 participants, consisting of 38 males and 82 females. Students’ writing capabilities using the generative tools were evaluated across different dimensions as part of a broader research framework. The findings emphasize the significance of AI advancements in enhancing students' writing skills while preserving key human-centered teaching dynamics. Future studies should be conducted with a bigger sample size with more universities using more advanced statistics. Qualitative and combined designs should also be taken into consideration.

Systematic Review of Generative Modelling Tools and Utility Metrics for Fully Synthetic Tabular Data

Sharing data with third parties is essential for advancing science, but it is becoming more and more difficult with the rise of data protection regulations, ethical restrictions, and growing fear of misuse. Fully synthetic data, which transcends anonymisation, may be the key to unlocking valuable untapped insights stored away in secured data vaults. This review examines current synthetic data generation methods and their utility measurement. We found that more traditional generative models such as Classification and Regression Tree models alongside Bayesian Networks remain highly relevant and are still capable of surpassing deep learning alternatives like Generative Adversarial Networks. However, our findings also display the same lack of agreement on metrics for evaluation, uncovered in earlier reviews, posing a persistent obstacle to advancing the field. We propose a tool for evaluating the utility of synthetic data and illustrate how it can be applied to three synthetic data generation models. By streamlining evaluation and promoting agreement on metrics, researchers can explore novel methods and generate compelling results that will convince data curators and lawmakers to embrace synthetic data. Our review emphasises the potential of synthetic data and highlights the need for greater collaboration and standardisation to unlock its full potential.

Comparative Analysis of Generative AI Tools in Making Marketing Copywriting in the Era of New Media

In the background of the development of new media, generative Artificial Intelligence (AI) tools have become powerful assets to create marketing copywriting, which provides efficiency and innovation ability for the generation of content. This study analyzes the performance of 4 types of AI tools ChatGPT4o, Image-TO-Prompt-AI(Poe), Tiangong3.0 and Doubao when producing and promoting copywriting. The study discussed the performance of these AI to generate the text which can cause emotional resonance, capture key marketing elements and show special language style. By applying the BROKE instruction framework, this study ensures that the evaluation method of AI-generated content is consistent and structured. Research results reveal significant differences in content quality through user feedback and questionnaire collection, among them, ChatGPT4o is leading in overall attractiveness and TianGong3.0 has an outstanding performance on the unique perspective and logical structure. This study not only provides insights into the current capabilities of AI in content creation, but also lays the foundation for the future development of AI and artificial collaboration in the new media field.

A Review of How AI Affects Education in Marketing Communications in College

The integration of artificial intelligence (AI) in higher education (HE) has expanded significantly in recent years. Colleges and other educational institutions are gradually adopting new technologies in teaching and learning. The development of technology and large model algorithms have greatly enhanced AI-generated content's capabilities, positioning AI as promising generative tools that add convenience and potential to education. While much research has focused on AI's role in disciplines like Engineering and Computer Science, its impact on Marketing Communications (Marcom) education remains under-explored. This review investigates how AI tools enhance Marcom education by simulating real-world marketing environments because Marcom is a career-oriented discipline. This review discusses AI in HE and its applications within the Marcom industry and then bridges the gap between HE and the industry by exploring Marcom education's current situation and future opportunities with AI assistants. The findings highlight the importance of equipping both educators and students with AI competencies. Educators can use AI to improve the teaching process, and they also play an essential role in monitoring students' AI usage to ensure their learning quality. From the students' aspects, mastering AI tools can help them be well-prepared for the evolving Marcom industry.

Analyzing the justification for using generative AI technology to generate judgments based on the virtue jurisprudence theory

ABSTRACT. This paper responds to the question of whether judgements generated by judges using ChatGPT can be directly adopted. It posits that it is unjust for judges to rely on and directly adopt ChatGPT-generated judgements based on virtue jurisprudence theory. This paper innovatively applies case-based empirical analysis and is the first to use virtue jurisprudence approach to analyse the question and support its argument. The first section reveals the use of generative AI-based tools in judicial practice and the existence of erroneous judgements generated by ChatGPT through empirical research. The second section contends that generative AI, while mimicking virtuous behaviour, cannot produce justified beliefs according to virtue jurisprudence theory. Moreover, using AI as a decision-making entity could undermine the protection of human rights and interests. The paper concludes by suggesting normative and technical approaches for how judges should use generative AI-based tools to develop their virtues.

Exploring GPT‐4.0's efficiency in handling paediatric appendicitis questions

AbstractObjectiveTo explore the potential and accuracy of the generative dialogue artificial intelligence tool GPT‐4.0 in answering questions related to paediatric emergency appendicitis.MethodsA cross‐sectional observational study design was used. We collected 134 appendicitis‐related questions from authoritative websites, such as Mayo Medical and APSA, covering all aspects of appendicitis, including simple and complex questions. These questions were answered by GPT‐4.0, and then evaluated by three paediatric surgical experts using a quality score ranging from 0 to 5. The answers were generated by GPT‐4.0 and then similarly evaluated by three experts for accuracy.ResultsWe found that GPT‐4.0 could achieve a high accuracy rate on simple questions with a quality score of 4.65 (standard deviation 0.51). For complex questions, the average score was 3.77 (standard deviation 0.68), and there was a significant difference between the two (P < .05). On clinical questions, the accuracy score of GPT‐4.0 was 4.00 (standard deviation 0.21). When answering actual questions from families of children with appendicitis, the accuracy score was 4.12 (standard deviation 0.59). Its accuracy lies between simple questions and complex questions, and it can basically meet the accuracy requirements of clinical questions. It's worth noting that GPT‐4.0 demonstrated empathy in answering some questions, which might further enhance patient satisfaction.ConclusionGPT‐4.0 showed its potential and accuracy in handling paediatric appendicitis questions, especially in simple and clinical questions. However, improvements are still needed in handling complex questions and updating information. Despite the limitations, this model is expected to improve the quality of medical services and enhance patient satisfaction.

Analysis and Prediction of Chymotrypsin Substrate Preferences through Large Data Acquisition with Target-Free mRNA Display.

Oral delivery of peptide therapeutics is limited by degradation by gut proteases like chymotrypsin. Existing databases of peptidases are limited in size and do not enable systematic analyses of protease substrate preferences, especially for non-natural amino acids. Thus, stability optimization of hit compounds is time and resource intensive. To accelerate the stability optimization of peptide ligands, we generated large datasets of chymotrypsin-resistant peptides via mRNA display to create a predictive model for chymotrypsin-resistant sequences. Through analysis of enriched motifs, we recapitulate known chymotrypsin cleavage sites, reveal positionally dependent effects of monomers on peptide cleavage, and report previously unidentified protective and destabilizing residues. We then developed a machine-learning-based model predicting peptide resistance to chymotrypsin cleavage and validated both model performance and the NGS experimental data by measuring chymotrypsin half-lives for a subset of peptides. Finally, we simulated stability predictions on non-natural amino acids through a leucine hold-out model and observed robust performance. Overall, we demonstrate the utility of mRNA display as a tool for big data generation and show that pairing mRNA display with machine learning yields valuable predictions for chymotrypsin cleavage. Expansion of this workflow to additional proteases could provide complementary predictive models that focus future peptide drug discovery efforts.

The use of Artificial Intelligence in the restaurant business

The study aimed to explore how artificial intelligence (AI) is applied in the restaurant industry, particularly in enhancing operational efficiency, personalizing customer interactions, and predicting demand. Using qualitative methods, interviews were conducted with industry experts in Odesa, Ukraine, including professionals from major chains and a case study with a Ukrainian chef. Findings revealed that global brands like McDonald’s, Starbucks, and Marriott use AI for demand forecasting, offer personalization, and streamlining communication. Chef Yevhen Klopotenko’s use of AI to prepare a gourmet dinner showcased its innovative potential in cuisine. The research enriches the existing literature on AI in hospitality with practical examples, offering valuable insights for industry professionals. It highlights the significance of AI in predicting demand and automating customer interactions to boost efficiency. The originality of the study lies in its analysis of AI applications and its comprehensive list of AI tools for content generation. However, the qualitative focus and regional scope limit generalizability, suggesting future research could broaden to include other regions and quantitative methods.

Exploring Students’ Multimodal Representations of Ideas About Epistemic Reading of Scientific Texts in Generative AI Tools

Exploring Students’ Multimodal Representations of Ideas About Epistemic Reading of Scientific Texts in Generative AI Tools

A comparative analysis of text-to-image generative AI models in scientific contexts: a case study on nuclear power

In this work, we propose and assess the potential of generative artificial intelligence (AI) as a tool for facilitating public engagement around potential clean energy sources. Such an application could increase energy literacy—an awareness of low-carbon energy sources among the public therefore leading to increased participation in decision-making about the future of energy systems. We explore the use of generative AI to communicate technical information about low-carbon energy sources to the general public, specifically in the realm of nuclear energy. We explored 20 AI-powered text-to-image generators and compared their individual performances on general and scientific nuclear-related prompts. Of these models, DALL-E, DreamStudio, and Craiyon demonstrated promising performance in generating relevant images from general-level text related to nuclear topics. However, these models fall short in three crucial ways: (1) they fail to accurately represent technical details of energy systems; (2) they reproduce existing biases surrounding gender and work in the energy sector; and (3) they fail to accurately represent indigenous landscapes—which have historically been sites of resource extraction and waste deposition for energy industries. This work is performed to motivate the development of specialized generative tools to improve energy literacy and effectively engage the public with low-carbon energy sources.

Emulation of f(R) modified gravity from ΛCDM using conditional GANs

Abstract A major aim of cosmological surveys is to test deviations from the standard ΛCDM model, but the full scientific value of these surveys will only be realised through efficient simulation methods that keep up with the increasing volume and precision of observational data. N-body simulations of modified gravity (MG) theories are computationally expensive since highly non-linear equations must be solved. This represents a significant bottleneck in the path to reach the data volume and resolution attained by equivalent ΛCDM simulations. We develop a field-level neural network-based emulator that generates density and velocity divergence fields under the f(R) gravity MG model from the corresponding ΛCDM simulated fields. Using attention mechanisms and a complementary frequency-based loss function, our model is able to learn this intricate mapping. We use the idea of latent space extrapolation to generalise our emulator to f(R) models with differing field strengths. The predictions of our emulator agree with the f(R) simulations to within 5% for matter density and to within 10% for velocity divergence power spectra up to k ∼ 2 h Mpc−1. But for a few select cases, higher-order statistics are reproduced with ≲10% agreement. Latent extrapolation allows our emulator to generalise to different parameterisations of the f(R) model without explicitly training on those variants. Given a ΛCDM simulation, the GPU-based emulator can reproduce the equivalent f(R) realisation ∼600 times faster than full N-body simulations. This lays the foundations for a valuable tool for realistic yet rapid mock field generation and robust cosmological analyses.

Generative AI in Writing Workshops: A Path to AI Literacy

The widespread use of generative AI tools which can support or even take over several part of the writing process has sparked many discussions about integrity, AI literacy and changes to academic writing processes. This paper explores the impact of generative artificial intelligence (AI) tools on the academic writing pro-cess, drawing on data from a writing workshop and interviews with university students of a university teacher college in Austria. Despite the widespread assumption that generative AI, such as ChatGPT, is widely used by students to support their academic tasks, initial findings suggest a notable gap in participants' experience and understanding of these technologies. This discrepancy highlights the critical need for AI literacy and underscores the importance of familiarity with the potential, challenges and risks associated with generative AI to ensure its ethical and effective use. Through reflective discussions and feedback from work-shop participants, this study shows a differentiated perspective on the role of generative AI in academic writing, illustrating its value in generating ideas and overcoming writer's block, as well as its limitations due to the indispensable nature of human involvement in critical writing tasks.

Can diffusion models capture extreme event statistics?

For many important problems it is essential to be able to accurately quantify the statistics of extremes for specific quantities of interest, such as extreme atmospheric weather events or ocean-related quantities. While there are many classical approaches to perform such modeling tasks, recent interest has been increasing in the usage of generative models trained on available data. Despite the sporadic success of such methods, it is not clear for what systems or datasets a system-agnostic generative AI tool is capable of generating previously ‘unseen’ extreme events in a manner that accurately extrapolates the tails for the observable of interest. Here, we propose an apriori criterion, which based on the geometry of the training dataset, it can predict whether a generative AI tool will be able to extrapolate the tails, i.e. generate previously unseen extreme events. The idea is to quantify whether existing extreme events lie in the interior of the dataset or its boundary. In the former case it is shown that generative AI tools can work in an ‘interpolation’ mode and generate new extreme events. On the other hand, if the topology of the dataset is such that extremes live in the boundary of the domain then the generative AI algorithm needs to operate in an extrapolation mode which does not lead to accurate results. We illustrate our findings on a specific class of Diffusion Models (DMs) called Denoising Diffusion Probabilistic Models (DDPMs) and we test on three datasets, a simple on-hyperball dataset following a Weibull distribution for the radii of the data points of dimensionality 2⋅103, a dataset sampled from the so-called Majda–McLaughlin–Tabak Wave Model (MMT), of dimensionality 8.1⋅103 and a dataset consisting of Lagrangian turbulence trajectories, of dimensionality 2⋅103.

Enhancing Assessment and Feedback in Game Design Programs

The integration of generative AI tools in game design education offers promising ways to streamline the grading, assessment, and feedback processes that are typically labor-intensive. In game design programs, faculty often deal with varied file formats, including 3D models, executable prototypes, videos, and complex game design documents. Traditional methods of assessment and feedback, primarily text-based, struggle to provide timely and actionable insights for students. Furthermore, only a small percentage of top students consistently review and apply feedback, leading to inefficiencies. This article explores how generative AI tools can augment these processes by automating aspects of grading, generating more personalized and meaningful feedback, and addressing the time-intensive nature of reviewing diverse file formats. Key strategies are discussed, including the use of rubrics tailored for AI-based assessment, automated prompts for narrative-driven assignments, and the application of AI in reviewing complex project builds. The objective is to create more time for faculty to engage in live mentoring and hands-on learning activities, which research shows to be more effective. Practical examples of various game design assignments, including build reviews and document evaluations, are provided to illustrate these new approaches. This shift promises to enhance student engagement and improve learning outcomes.

Facilitating AADL Model Processing and Analysis with OSATE-DIM

The Architecture Analysis and Design Language (AADL) is a rich component-based language for modelling embedded systems. To ease processing AADL models, OSATE, the reference tool for AADL, provides the 'instance' model derived from base 'declarative' model/s. An instance model represents the operational view of a declarative model in a simple object tree where information is flattened (with no component extensions / refinements) so that tools can easily analyze the model. Note that information is lost in instantiation. Since the instance model is a (un-symmetric) 'view' of the declarative model, the capability to directly modify the instance model requires a solution to the view-update problem. We demonstrate the OSATE Declarative-Instance Mapping Tool (OSATEDIM) to perform incremental deinstantiation in AADL. OSATE-DIM significantly eases the development of AADL model processing tools for analysis and code generation.

An investigation on the collapse pressure prediction of subsea pipelines with realistic corrosion defects

Subsea pipelines are widely used for oil and gas transportation, but they are susceptible to failure due to corrosion. Corrosion in pipelines typically results in defects of varying depths and complex shapes. The computational simulation through the finite element (FE) method is one of the most efficient and accurate tools for assessing the integrity of corroded subsea pipes by allowing the simulation of the nonlinear behavior and assessing the hydrostatic collapse. This paper evaluates the collapse response of subsea pipelines with realistic corrosion defects. The study uses the PIPEFLAW system - developed by the PADMEC (High-Performance Processing on Computational Mechanics) research group from UFPE - which integrates reliable tools for automatic FE model generation and performs nonlinear analyses. After validation based on the experimental tests available in the literature, the effect of realistic corrosion defects on the collapse pressure of subsea pipelines is analyzed in detail. The results are compared to results obtained using a semi-empirical method and numerical results obtained considering idealized FE models, i.e., constant-depth and elliptical-shaped corrosion defects. It is observed that the non-uniformity of defects is an important factor affecting the collapse response of pipes. In addition, the idealized geometry of the corrosion defects, especially when considering constant depth, leads to an excessively conservative prediction of the hydrostatic collapse.

Reconfigurable and Scalable Honeynet for Cyber-physical Systems

Industrial Control Systems (ICS) constitute the backbone of contemporary industrial operations, ranging from modest heating, ventilation, and air conditioning systems to expansive national power grids. Given their pivotal role in critical infrastructure, there has been a concerted effort to enhance security measures and deepen our comprehension of potential cyber threats within this domain. To address these challenges, numerous implementations of Honeypots and Honeynets intended to detect and understand attacks have been employed for ICS. This approach diverges from conventional methods by focusing on making a scalable and reconfigurable honeynet for cyber-physical systems. It will also automatically generate attacks on the honeynet to test and validate it. With the development of a scalable and reconfigurable Honeynet and automatic attack generation tools, it is also expected that the system will serve as a basis for producing datasets for training algorithms for detecting and classifying attacks in cyberphysical honeynets.

Students’ Writing Ability By Using AI Generative Tools; Diffit, Brisk, Mendeley

This study examines the impact of three AI generative tools—Diffit, Brisk, and Mendeley—on university students’ writing perceptions and motivation. A mixed-methods approach combined quantitative pretest and posttest assessments with qualitative insights from interviews and classroom observations. The participants, 27 students from Universitas Ekasakti Padang, were purposively sampled and randomly assigned to groups based on the AI tool used. Quantitative results revealed that the Diffit group achieved self-efficacy improvements from 80% (initial) to 90% (post-intervention), with a task value of 60%. The Brisk group showed consistently high self-efficacy (90%) and task value (90%), achieving an 80% writing proficiency score. In contrast, the Mendeley group recorded self-efficacy scores of 100% initially, dropping to 60%, but maintained a writing proficiency score of 80% due to a task value of 80%. Findings highlight self-efficacy as the most critical factor influencing writing improvement, supported by task value, intrinsic motivation, and perceived usefulness. Recommendations include integrating AI tools to build students’ confidence, aligning tasks with personal goals, and fostering intrinsic motivation. This study demonstrates the effectiveness of AI tools in enhancing writing proficiency by addressing students' cognitive, emotional, and motivational needs.

Exploring spatial reasoning performances of CNN on linear layout dataset

Abstract Spatial reasoning, a fundamental aspect of human intelligence, is essential for machine learning models to understand and interpret object relationships. It is crucial for numerous real-world applications, ranging from autonomous navigation to urban planning. The lack of comprehensive datasets limits the development and evaluation of models that can effectively handle spatial reasoning tasks. Existing datasets often contain complex spatial reasoning problems with overlapping spatial relationships, making it challenging to diagnose specific aspects that a model struggles with. We address this gap by introducing a new dataset of linear layouts. This dataset is systematically designed to exhibit a range of spatial relations and complexity levels. Analyzing spatial reasoning through linear layout generation offers a more structured and manageable approach to understanding how models learn and interpret spatial relationships. Linear layout generation has broad applicability and is of fundamental importance in design and optimization. To benchmark dataset, we develop LinLayCNN, a generic data-driven method that applies shallow, one-dimensional convolutional neural network (CNN), to generate linear layouts in an iterative process. Experimental results reveal that LinLayCNN can effectively solve fundamental spatial challenges even with the relatively small size of the training set. It is capable of precise object placement, making it a robust tool for linear layout generation. Current layout generation methods focus on domain-specific solutions and often fail to maintain the precision needed for technical domains, such as accurate sizing, and object counting. They also require a substantial amount of data to function effectively. LinLayCNN overcame these issues. This study further clarifies CNNs’ capabilities in spatial reasoning, highlight their potential to advance the field of layout generation. As a result, our approach establishes a clear benchmark for evaluating spatial reasoning and aids in development of models that can more effectively understand and reason about space.

An empirical study of factors influencing usage intention for generative artificial intelligence products: A case study of China

The rapid advancement of generative artificial intelligence technologies has sparked widespread interest, yet understanding of user adoption patterns for these tools remains limited. While the unified theory of acceptance and use of technology model has been widely applied to various technologies, its applicability to generative artificial intelligence products, which present unique challenges and opportunities, has not been thoroughly explored. This gap in knowledge hinders the development of effective strategies for promoting user acceptance and optimising the design of generative artificial intelligence tools. This study extends the unified theory of acceptance and use of technology model to investigate the factors influencing usage intention for generative artificial intelligence products. The study incorporates additional variables such as digital literacy, perceived risk and perceived trust to provide a more comprehensive framework. Using structural equation modelling to analyse survey data, the study found that effort expectancy, performance expectancy, social influence, and perceived trust significantly and positively impact usage intention. In addition, digital literacy indirectly enhances usage intention through effort expectancy, while perceived risk negatively influences usage intention through reduced trust. Notably, facilitating conditions did not exhibit a significant effect on usage intention. These findings offer valuable insights for developers and researchers in the field of generative artificial intelligence, highlighting the importance of user-friendly design, performance optimisation, and trust-building measures. By identifying key factors that drive user adoption, this study contributes to a more nuanced understanding of technology acceptance in the context of advanced artificial intelligence systems, paving the way for more effective development and implementation strategies in this rapidly evolving field.