Domain-specific Language Research Articles

Using Music Features for Managing Revisions and Variants of Musical Scores

Abstract Music engravers nowadays use music notation software to create scores for musical works. As is common in any creative process, many different versions of digital artifacts are created—e.g., to manage editions of the same musical work, editorial markups, or the history and genesis of compositions. In the field of software engineering, researchers have proposed the use of features, i.e., user-visible aspects of systems, to manage both revisions and variants of source code and other software artifacts. Feature-based version control systems establish and maintain feature-to-code mappings defining which parts of the artifacts realize particular features, e.g., which code implements a specific function. These mappings can then be used to generate new variants based on the artifacts by selecting the desired features. Our article provides an in-depth study on feature-based version control for music notation. Our automated approach adopts the domain-specific language (DSL) LilyPond and the feature-oriented version control system ECCO. Existing studies show that features in musical scores are often fine-grained and affect only small parts of an artifact, are scattered across noncontiguous locations in the artifact, and highly interact with each other. Such properties have strong implications for the usefulness of versioning tools. Our experiment investigates two factors related to the correctness of output from feature-based version control systems when used for symbolic music notation. We demonstrate the incremental refinement of feature-to-artifact mappings when committing DSL code. We further study the impact of the order of feature interactions on the correctness of the automatically generated music artifacts. We find that a larger feature interaction threshold produces only marginally more correct results, but fixing and recommitting incorrect variants has a more powerful effect. Our results further show that considering DSL specifics is important for versioning fine-grained and scattered features.

Necessity and impact of specialization of large foundation model for medical segmentation tasks.

Large foundation models, such as the Segment Anything Model (SAM), have shown remarkable performance in image segmentation tasks. However, the optimal approach to achieve true utility of these models for domain-specific applications, such as medical image segmentation, remains an open question. Recent studies have released a medical version of the foundation model MedSAM by training on vast medical data, who promised SOTA medical segmentation. Independent community inspection and dissection is needed. Foundation models are developed for general purposes. On the other hand, stable delivery of reliable performance is key to clinical utility. This study aims at elucidating the potential advantage and limitations of landing the foundation models in clinical use by assessing the performance of off-the-shelf medical foundation model MedSAM for the segmentation of anatomical structures in pelvic MR images. We also explore the simple remedies by evaluating the dependency on prompting scheme. Finally, we demonstrate the need and performance gain of further specialized fine-tuning. MedSAM and its lightweight version LiteMedSAM were evaluated out-of-the-box on a public MR dataset consisting of 589 pelvic images split 80:20 for training and testing. An nnU-Net model was trained from scratch to serve as a benchmark and to provide bounding box prompts for MedSAM. MedSAM was evaluated using different quality bounding boxes, those derived from ground truth labels, those derived from nnU-Net, and those derived from the former two but with 5-pixel isometric expansion. Lastly, LiteMedSAM was refined on the training set and reevaluated on this task. Out-of-the-box MedSAM and LiteMedSAM both performed poorly across the structure set, especially for disjoint or non-convex structures. Varying prompt with different bounding box inputs had minimal effect. For example, the mean Dice score and mean Hausdorff distances (in mm) for obturator internus using MedSAM and LiteMedSAM were {0.251±0.110, 0.101±0.079} and {34.142±5.196, 33.688±5.306}, respectively. Fine-tuning of LiteMedSAM led to significant performance gain, improving Dice score and Hausdorff distance for the obturator internus to 0.864±0.123 and 5.022±10.684, on par with nnU-Net with no significant difference in evaluation of most structures. All segmentation structures benefited significantly from specialized refinement, at varying improvement margin. While our study alludes to the potential of deep learning models like MedSAM and LiteMedSAM for medical segmentation, it highlights the need for specialized refinement and adjudication. Off-the-shelf use of such large foundation models is highly likely to be suboptimal, and specialized fine-tuning is often necessary to achieve clinical desired accuracy and stability.

Exploring the Potential of Tunnel Field-Effect Transistors in Biomedical Devices: A Comprehensive Survey

Implantable medical devices have revolutionized healthcare by providing continuous monitoring and therapeutic interventions within the human body. The overall performance of these devices is heavily dependent on the effectiveness and efficiency of integrated electronic components, such as transistors. This survey explores the growing utilization of Tunnel Field-Effect Transistors (TFETs) in implantable devices, shedding light on their advantages and challenges in this specific application domain. TFETs, characterized by their unique tunnelling mechanism, offer several advantages that make them well suited for implantable devices. Their low-power consumption, reduced leakage current, and improved subthreshold swing make them ideal for energy-efficient, long-lasting operation within the constrained power budgets of implantable devices. Furthermore, TFETs exhibit excellent performance at low supply voltages, ensuring minimal tissue damage and heat generation. This survey also addresses the challenges associated with TFETs in implantable devices, including fabrication complexities, variability, and compatibility with existing technologies. It highlights recent advancements in TFET technology that address these concerns, such as improved material choices and manufacturing techniques. As technology continues to advance, TFETs are poised to play a pivotal role in improving the overall efficiency and reliability of life-saving technologies and medical devices. Overall, this survey provides valuable insights into the current state and future prospects of TFETs in implantable devices, guiding researchers and engineers in harnessing their potential for the benefit of healthcare and patient well-being.

A Proactive Decoy Selection Scheme for Cyber Deception using MITRE ATT&CK

Cyber deception allows compensating the late response of defenders countermeasures to the ever evolving tactics, techniques, and procedures (TTPs) of attackers. This proactive defense strategy employs decoys resembling legitimate system components to lure stealthy attackers within the defender environment, slowing and/or denying the accomplishment of their goals. In this regard, the selection of decoys that can expose the techniques used by malicious users plays a central role to incentivize their engagement. However, this is a difficult task to achieve in practice, since it requires an accurate and realistic modeling of the attacker capabilities and his possible targets. In this work, we tackle this challenge and we design a decoy selection scheme that is supported by an adversarial modeling based on empirical observation of real-world attackers. We take advantage of a domain-specific threat modeling language using MITRE ATT&CK© framework as source of attacker TTPs targeting enterprise systems. In detail, we extract the information about the execution preconditions of each technique as well as its possible effects on the environment to generate attack graphs modeling the adversary capabilities. Based on this, we formulate a graph partition problem that minimizes the number of decoys detecting a corresponding number of techniques employed in various attack paths directed to specific targets. We compare our optimization-based decoy selection approach against several benchmark schemes that ignore the preconditions between the various attack steps. Results reveal that the proposed scheme provides the highest interception rate of attack paths using the lowest amount of decoys.

When Your Infrastructure Is a Buggy Program: Understanding Faults in Infrastructure as Code Ecosystems

Modern applications have become increasingly complex and their manual installation and configuration is no longer practical. Instead, IT organizations heavily rely on Infrastructure as Code (IaC) technologies, to automate the provisioning, configuration, and maintenance of computing infrastructures and systems. IaC systems typically offer declarative, domain-specific languages (DSLs) that allow system administrators and developers to write high-level programs that specify the desired state of their infrastructure in a reliable, predictable, and documented fashion. Just like traditional programs, IaC software is not immune to faults, with issues ranging from deployment failures to critical misconfigurations that often impact production systems used by millions of end users. Surprisingly, despite its crucial role in global infrastructure management, the tooling and techniques for ensuring IaC reliability still have room for improvement. In this work, we conduct a comprehensive analysis of 360 bugs identified in IaC software within prominent IaC ecosystems including Ansible, Puppet, and Chef. Our work is the first in-depth exploration of bug characteristics in these widely-used IaC environments. Through our analysis we aim to understand: (1) how these bugs manifest, (2) their underlying root causes, (3) their reproduction requirements in terms of system state (e.g., operating system versions) or input characteristics, and (4) how these bugs are fixed. Based on our findings, we evaluate the state-of-the-art techniques for IaC reliability, identify their limitations, and provide a set of recommendations for future research. We believe that our study helps researchers to (1) better understand the complexity and peculiarities of IaC software, and (2) develop advanced tooling for more reliable and robust system configurations.

Semantics Lifting for Syntactic Sugar

Syntactic sugar plays a crucial role in engineering programming languages. It offers convenient syntax and higher-level of abstractions, as witnessed by its pervasive use in both general-purpose and domain-specific contexts. Unfortunately, the traditional approach of translating programs containing syntactic sugars into the host language can lead to abstraction leakage, breaking the promise of convenience and hindering program comprehension. To address this challenge, we introduce the idea of semantics lifting that aims to statically derive self-contained evaluation rules for syntactic sugars. More specifically, we propose a semantics-lifting framework that consists of (i) a general algorithm for deriving host-independent semantics of syntactic sugars from the semantics of the host language and the desugaring rules, (ii) a formulation of the correctness and abstraction properties for a lifted semantics, and (iii) a systematic investigation of sufficient conditions that ensure a lifted semantics is provably correct and abstract. To evaluate our semantics-lifting framework, we have implemented a system named Osazone and conducted several case studies, demonstrating that our approach is flexible, effective, and practical for implementing domain-specific languages.

Serious Games for Vocational Training

The research project ‘SG4BB’ (Serious Games for Vocational Education and Training; abbr.: VET) has developed an integrated platform for the description, search, retrieval, integration, and utilization of educational games. This paper summarizes project outcomes, including platform concepts, software components, and – as focus – practical insights derived from a case study involving the educational game ‘Corrugated’ as simulation and training environment for service technicians. SG4BB has followed a user-centered design in an interdisciplinary project team: Two learning providers specified their requirements for VET in their specific application domains. Based on those requirements, game developers, learning solution providers and researchers for VET have conceptualized and prototypically implemented both an integrated learning platform (SG4BB platform) and two case studies of Serious Games for VET: ‘Corrugated’ for service technicians and an IT security game. The platform follows a process pipeline: For search and retrieval of educational games for VET, an application profile for VET based on the standardized ‘Serious Games Metadata Format’ (DIN/SPEC 91380) has been elaborated. This format builds the semantic basis for the metadata-based catalog system ‘Serious Games Information Center’ (SG-IC) with filter functionality for VET. Learning providers and developers can use the SG-IC portal to describe and promote their educational games, enabling users to identify suitable games for their learning needs and integrate them via learning infrastructure. Educational games can interact with the backend (Learning Management System and Learning Record Store) through a middleware based on the xAPI standard, allowing for personalized gameplay and data collection for game-based learning analytics. The final evaluation of the SG4BB project focused on the utilization of the educational game ‘Corrugated’, targeting problem-solving skills for service technicians in the corrugated cardboard industry. Data from 26 participants playing the game for 60 minutes, along with problem-solving tests and user experience feedback, were analyzed to validate game-based assessment and to assess learning impact. Initial results reveal insights into specific game missions, playtimes, and success rates, indicating that participant behavior during gameplay influenced perceived learning progress, leading to varied learning paths. This paper provides valuable insights and technical information for VET practitioners on using educational games for training. Game interactions and learning outcomes can be monitored via a dashboard within the learning infrastructure, offering visualizations for user behavior (during play) and (learning) progress. Keywords: serious games, vocational education and training, metadata, middleware, case study and evaluation.

HERALD: A domain-specific query language for longitudinal health data analytics

BackgroundLarge-scale health data has significant potential for research and innovation, especially with longitudinal data offering insights into prevention, disease progression, and treatment effects. Yet, analyzing this data type is complex, as data points are repeatedly documented along the timeline. As a consequence, extracting cross-sectional tabular data suitable for statistical analysis and machine learning can be challenging for medical researchers and data scientists alike, with existing tools lacking balance between ease-of-use and comprehensiveness. ObjectiveThis paper introduces HERALD, a novel domain-specific query language designed to support the transformation of longitudinal health data into cross-sectional tables. We describe the basic concepts, the query syntax, a graphical user interface for constructing and executing HERALD queries, as well as an integration into Informatics for Integrating Biology and the Bedside (i2b2). MethodsThe syntax of HERALD mimics natural language and supports different query types for selection, aggregation, analysis of relationships, and searching for data points based on filter expressions and temporal constraints. Using a hierarchical concept model, queries are executed individually for the data of each patient, while constructing tabular output. HERALD is closed, meaning that queries process data points and generate data points. Queries can refer to data points that have been produced by previous queries, providing a simple, but powerful nesting mechanism. ResultsThe open-source implementation consists of a HERALD query parser, an execution engine, as well as a web-based user interface for query construction and statistical analysis. The implementation can be deployed as a standalone component and integrated into self-service data analytics environments like i2b2 as a plugin. HERALD can be valuable tool for data scientists and machine learning experts, as it simplifies the process of transforming longitudinal health data into tables and data matrices. ConclusionThe construction of cross-sectional tables from longitudinal data can be supported through dedicated query languages that strike a reasonable balance between language complexity and transformation capabilities.

OpenSBLI v3.0: High-Fidelity Multi-Block Transonic Aerofoil CFD Simulations using Domain Specific Languages on GPUs

OpenSBLI is an automatic code-generation framework for compressible Computational Fluid Dynamics (CFD) simulations on heterogeneous computing architectures (previous release: Lusher et al, CPC (2021)). OpenSBLI is coupled to the Oxford Parallel Structured (OPS) Domain Specific Language (DSL), which uses source-to-source translation to enable parallel execution of the code on large-scale supercomputers, including multi-GPU clusters. To date, OpenSBLI has largely been applied to compressible turbulence and shock-wave/boundary-layer interactions on very simple geometries comprised of single mesh blocks with essentially orthogonal grid lines. OpenSBLI has been extended in this new release to target strongly curvilinear cases, including transonic aerofoils using multi-block grids. In addition to multi-block mesh support, more efficient numerical shock-capturing methods and filters have been added to the codebase. Improvements to post-processing, reduced-dimension data output, and coupling to a modal decomposition library are also included. A set of validation cases are presented to showcase the new code features. Furthermore, state-of-the-art wide-span transonic aerofoil simulations on up to N=2.5×109 grid points demonstrate that wider aspect ratios can alter buffet predictions and increase the regularity of the low-frequency shock oscillations by accommodating fully-developed trailing edge flow separation. Spectral Proper Orthogonal Decomposition (SPOD) analysis showed that overly-narrow aerofoil simulations contain additional domain-dependent energy content at a Strouhal number of St≈3 associated with wake modes.

A method for evaluating deep generative models of images for hallucinations in high-order spatial context

Deep generative models (DGMs) have the potential to revolutionize diagnostic imaging. Generative adversarial networks (GANs) are one kind of DGM which are widely employed. The overarching problem with deploying any sort of DGM in mission-critical applications is a lack of adequate and/or automatic means of assessing the domain-specific quality of generated images. In this work, we demonstrate several objective and human-interpretable tests of images output by two popular DGMs. These tests serve two goals: (i) ruling out DGMs for downstream, domain-specific applications, and (ii) quantifying hallucinations in the expected spatial context in DGM-generated images. The designed datasets are made public and the proposed tests could also serve as benchmarks and aid the prototyping of emerging DGMs. Although these tests are demonstrated on GANs, they can be employed as a benchmark for evaluating any DGM. Specifically, we designed several stochastic context models (SCMs) of distinct image features that can be recovered after generation by a trained DGM. Together, these SCMs encode features as per-image constraints in prevalence, position, intensity, and/or texture. Several of these features are high-order, algorithmic pixel-arrangement rules which are not readily expressed in covariance matrices. We designed and validated statistical classifiers to detect specific effects of the known arrangement rules. We then tested the rates at which two different DGMs correctly reproduced the feature context under a variety of training scenarios, and degrees of feature-class similarity. We found that ensembles of generated images can appear largely accurate visually, and show high accuracy in ensemble measures, while not exhibiting the known spatial arrangements. The main conclusion is that SCMs can be engineered, and serve as benchmarks, to quantify numerous per image errors, i.e., hallucinations, that may not be captured in ensemble statistics but plausibly can affect subsequent use of the DGM-generated images.

Self-guided Knowledge-Injected Graph Neural Network for Alzheimer's Diseases.

Graph neural networks (GNNs) are proficient machine learning models in handling irregularly structured data. Nevertheless, their generic formulation falls short when applied to the analysis of brain connectomes in Alzheimer's Disease (AD), necessitating the incorporation of domain-specific knowledge to achieve optimal model performance. The integration of AD-related expertise into GNNs presents a significant challenge. Current methodologies reliant on manual design often demand substantial expertise from external domain specialists to guide the development of novel models, thereby consuming considerable time and resources. To mitigate the need for manual curation, this paper introduces a novel self-guided knowledge-infused multimodal GNN to autonomously integrate domain knowledge into the model development process. We propose to conceptualize existing domain knowledge as natural language, and devise a specialized multimodal GNN framework tailored to leverage this uncurated knowledge to direct the learning of the GNN submodule, thereby enhancing its efficacy and improving prediction interpretability. To assess the effectiveness of our framework, we compile a comprehensive literature dataset comprising recent peer-reviewed publications on AD. By integrating this literature dataset with several real-world AD datasets, our experimental results illustrate the effectiveness of the proposed method in extracting curated knowledge and offering explanations on graphs for domain-specific applications. Furthermore, our approach successfully utilizes the extracted information to enhance the performance of the GNN.

Explainable AI (XAI) in healthcare: Enhancing trust and transparency in critical decision-making

The integration of artificial intelligence (AI) in healthcare is revolutionizing diagnostic and treatment procedures, offering unprecedented accuracy and efficiency. However, the opacity of many advanced AI models, often described as "black boxes," creates challenges in adoption due to concerns around trust, transparency, and interpretability, particularly in high-stakes environments like healthcare. Explainable AI (XAI) addresses these concerns by providing a framework that not only achieves high performance but also offers insight into how decisions are made. This research explores the application of XAI techniques in healthcare, focusing on critical areas such as disease diagnostics, predictive analytics, and personalized treatment recommendations. The study will analyse various XAI methods, including model-agnostic approaches (LIME, SHAP), interpretable deep learning models, and domain-specific applications of XAI. It also evaluates the ethical implications, such as accountability and bias mitigation, and how XAI can foster collaboration between clinicians and AI systems. Ultimately, the goal is to create AI systems that are both powerful and trustworthy, promoting broader adoption in the healthcare sector while ensuring ethical and safe outcomes for patients.

Bridging MDE and AI: a systematic review of domain-specific languages and model-driven practices in AI software systems engineering

AbstractTechnical systems are becoming increasingly complex due to the increasing number of components, functions, and involvement of different disciplines. In this regard, model-driven engineering techniques and practices tame complexity during the development process by using models as primary artifacts. Modeling can be carried out through domain-specific languages whose implementation is supported by model-driven techniques. Today, the amount of data generated during product development is rapidly growing, leading to an increased need to leverage artificial intelligence algorithms. However, using these algorithms in practice can be difficult and time-consuming. Therefore, leveraging domain-specific languages and model-driven techniques for formulating AI algorithms or parts of them can reduce these complexities and be advantageous. This study aims to investigate the existing model-driven approaches relying on domain-specific languages in support of the engineering of AI software systems to sharpen future research further and define the current state of the art. We conducted a Systemic Literature Review (SLR), collecting papers from five major databases resulting in 1335 candidate studies, eventually retaining 18 primary studies. Each primary study will be evaluated and discussed with respect to the adoption of (1) MDE principles and practices and (2) the phases of AI development support aligned with the stages of the CRISP-DM methodology. The study’s findings show that language workbenches are of paramount importance in dealing with all aspects of modeling language development (metamodel, concrete syntax, and model transformation) and are leveraged to define domain-specific languages (DSL) explicitly addressing AI concerns. The most prominent AI-related concerns are training and modeling of the AI algorithm, while minor emphasis is given to the time-consuming preparation of the data sets. Early project phases that support interdisciplinary communication of requirements, such as the CRISP-DM Business Understanding phase, are rarely reflected. The study found that the use of MDE for AI is still in its early stages, and there is no single tool or method that is widely used. Additionally, current approaches tend to focus on specific stages of development rather than providing support for the entire development process. As a result, the study suggests several research directions to further improve the use of MDE for AI and to guide future research in this area.

ScEMB: Learning context representation of genes based on large-scale single-cell transcriptomics.

The rapid advancement of single-cell transcriptomic technologies has led to the curation of millions of cellular profiles, providing unprecedented insights into cellular heterogeneity across various tissues and developmental stages. This growing wealth of data presents an opportunity to uncover complex gene-gene relationships, yet also poses significant computational challenges. We present scEMB, a transformer-based deep learning model developed to capture context-aware gene embeddings from large-scale single-cell transcriptomics data. Trained on over 30 million single-cell transcriptomes, scEMB utilizes an innovative binning strategy that integrates data across multiple platforms, effectively preserving both gene expression hierarchies and cell-type specificity. In downstream tasks such as batch integration, clustering, and cell type annotation, scEMB demonstrates superior performance compared to existing models like scGPT and Geneformer. Notably, scEMB excels in silico correlation analysis, accurately predicting gene perturbation effects in CRISPR-edited datasets and microglia state transition, identifying a few known Alzheimer's disease (AD) risks genes in top gene list. Additionally, scEMB offers robust fine-tuning capabilities for domain-specific applications, making it a versatile tool for tackling diverse biological problems such as therapeutic target discovery and disease modeling. scEMB represents a powerful tool for extracting biologically meaningful insights from complex gene expression data. Its ability to model in silico perturbation effects and conduct correlation analyses in the embedding space highlights its potential to accelerate discoveries in precision medicine and therapeutic development.

Effectiveness of lipid-based nutrient supplementation during the first 1000 days of life for early childhood development: A community-based trial from Pakistan.

A community-based, cluster non-randomized controlled trial was conducted in Kurram district, Pakistan between January 2018 to December 2020. Age-appropriate lipid-based nutrient supplements and health education (sessions conducted in the households) were given to pregnant women and their born children (6-23 months) in the intervention arm (n = 40 clusters) versus health education only in the control arm (n = 40 clusters) to evaluate its effect on child development. The first and second developmental assessments were completed at ~24 months (n = 689) and ~32 months (n = 608), respectively, using the Caregiver-Reported Early Development Instrument Long form. The overall and domain-specific (motor, language, cognitive and socio-emotional) scores were computed with higher scores indicating better child development. Higher development scores, including overall (β = 0.40, 95% confidence interval [CI]: 0.14, 0.65; p = 0.002), cognitive (β = 0.27, 95% CI: 0.10, 0.45; p = 0.002), motor (β = 0.39, 95% CI: 0.22, 0.56; p < 0.001) and language (β = 0.33, 95% CI: 0.14, 0.51; p = 0.001) were reported for children who received the intervention compared to the control arm at first developmental assessment. However, the effect was not sustained after the discontinuation of the intervention. The LNS received by the mothers (during pregnancy and first 6 months after delivery) and by children during 6-23 months of age was beneficial for the children. The trial is registered in the International Standard Randomised Controlled Trial Number Registry (ID: ISRCTN94319790) on December 11, 2017.

PreparedLLM: effective pre-pretraining framework for domain-specific large language models

ABSTRACT The direct application of large language models (LLMs) to specific domain tasks frequently encounters challenges due to the scarcity of domain data, variations in domain semantics, and the complexity of domain knowledge. Further pretraining of advanced foundational models on extensive domain-specific corpora can infuse these models with domain-specific knowledge and enhancing their ability of solving domain-specific tasks. However, the development of most domain-specific models focuses primarily on collecting large-scale domain data, often overlooking the crucial optimization of the pre-pretraining stage, which significantly impacts both model performance and training efficiency. This paper introduces PRE -PretrAining FRamEwork for Domain-specific Large Language Models (PreparedLLM), a framework designed to enhance the pre-pretraining stage for domain specialization of LLMs. PreparedLLM employs advanced techniques in data recipe, data cleaning, vocabulary expansion, and embedding initialization. These techniques are implemented to optimize both the composition and quality of the training data, and to enhance understanding of domain terminologies and concepts, as well as improve token embedding initialization. Utilizing the geoscience domain as a case study, this paper applies PreparedLLM for the domain specialization of the Llama, a widely recognized general-purpose LLM. Experimental results demonstrate that PreparedLLM enhances model convergence speed, training speed, inference speed, the text volume of the context window, and overall performance in domain specialization. The utilization of PreparedLLM in developing domain-specific LLMs has significantly increased performance while reducing both time and resource investment. The case study provides valuable insights into the development of domain-specific LLMs.

Advancing Sustainable Cyber-Physical System Development with a Digital Twins and Language Engineering Approach: Smart Greenhouse Applications

In recent years, the integration of Internet of Things technologies in smart agriculture has become critical for sustainability and efficiency, to the extent that recent improvements have transformed greenhouse farming. This study investigated the complexity of IoT architecture in smart greenhouses by introducing a greenhouse language family (GreenH) that comprises three domain-specific languages designed to address various tasks in this domain. The purpose of this research was to streamline the creation, simulation, and monitoring of digital twins, an essential tool for optimizing greenhouse operations. A three-stage methodology was employed to develop the GreenH DSLs, a detailed metamodel for enhanced smart monitoring systems. Our approach used high-level metamodels and extended Backus–Naur form notation to define the DSL syntax and semantics. Through a comprehensive evaluation strategy and a selected language usability metrics, the expressiveness, consistency, readability, correctness, and scalability of the DSL were affirmed, and areas for usability improvement were highlighted. The findings suggest that GreenH languages hold significant potential for advancing digital twin modeling in smart agriculture. Future work should be aimed at refining usability and extending its application range. The anticipated integration with additional model-drive engineering and code generation tools will improve interoperability and contribute to digital transformation in the smart greenhouse domain and promote more sustainable food production systems.

Travel counsellors’ perspectives on AI assistance

Research on AI in the context of travel counselling has mainly focused on increasing the efficiency and level of automation of online travel bookings. However, AI solutions can also be used in the context of offline travel counselling, i.e., face-to-face counselling at a travel agency service desk. In this specific application domain, AI has the potential to change personal interaction between counsellors and clients in different ways. In order to design AI solutions to enrich personal interaction, research on the specific requirements of counsellors, customers, and other relevant stakeholders in the context of AI-assisted offline travel counselling is needed. Human-centered design (HCD) is a promising approach to investigate and incorporate these perspectives on the course of the AI design process. In this paper, we present our approach of applying HCD elements to deepen the understanding of the context of use, to specify user requirements, and to find potential design solutions for a real-world use case of AI assistance in offline travel counselling. In addition, we highlight the key findings and insights regarding travel counsellors’ perspectives and requirements on possible AI adoption, showing that they are primarily open to AI-based innovations, but also have some concerns about the devaluation of the counselling profession. Further, we critically discuss additional design elements that would be needed for human-centered AI design in an ambitious sense.Practical Relevance: The findings are of high practical relevance as they demonstrate how HCD principles can be applied to real-life use cases, which aim to introduce human-AI interaction in work domains. At the same time, our findings provide valuable insights for practitioners into the specific work environment of offline travel counsellors.

RenderKernel: High-level programming for real-time rendering systems

Real-time rendering applications leverage heterogeneous computing to optimize performance. However, software development across multiple devices presents challenges, including data layout inconsistencies, synchronization issues, resource management complexities, and architectural disparities. Additionally, the creation of such systems requires verbose and unsafe programming models. Recent developments in domain-specific and unified shading languages aim to mitigate these issues. Yet, current programming models primarily address data layout consistency, neglecting other persistent challenges.In this paper, we introduce RenderKernel, a programming model designed to simplify the development of real-time rendering systems. Recognizing the need for a high-level approach, RenderKernel addresses the specific challenges of real-time rendering, enabling development on heterogeneous systems as if they were homogeneous. This model allows for early detection and prevention of errors due to system heterogeneity at compile-time. Furthermore, RenderKernel enables the use of common programming patterns from homogeneous environments, freeing developers from the complexities of underlying heterogeneous systems. Developers can focus on coding unique application features, thereby enhancing productivity and reducing the cognitive load associated with real-time rendering system development.

Performance of trauma-trained large language models on surgical assessment questions: A new approach in resource identification

BackgroundLarge language models have successfully navigated simulated medical board examination questions. However, whether and how language models can be used in surgical education is less understood. Our study evaluates the efficacy of domain-specific large language models in curating study materials for surgical board style questions. MethodsWe developed EAST-GPT and ACS-GPT, custom large language models with domain-specific knowledge from published guidelines from the Eastern Association of the Surgery of Trauma and the American College of Surgeons Trauma Quality Programs. EAST-GPT, ACS-GPT, and an untrained GPT-4 performance were assessed trauma-related questions from Surgical Education and Self-Assessment Program (18th edition). Large language models were asked to choose answers and provide answer rationales. Rationales were assessed against an educational framework with 5 domains: accuracy, relevance, comprehensiveness, evidence-base, and clarity. ResultsNinety guidelines trained EAST-GPT and 10 trained ACS-GPT. All large language models were tested on 62 trauma questions. EAST-GPT correctly answered 76%, whereas ACS-GPT answered 68% correctly. Both models outperformed ChatGPT-4 (P < .05), which answered 45% correctly. For reasoning, EAST-GPT achieved the gratest mean scores across all 5 educational framework metrics. ACS-GPT scored lower than ChatGPT-4 in comprehensiveness and evidence-base; however, these differences were not statistically significant. ConclusionOur study presents a novel methodology in identifying test-preparation resources by training a large language model to answer board-style multiple choice questions. Both trained models outperformed ChatGPT-4, demonstrating its answers were accurate, relevant, and evidence-based. Potential implications of such AI integration into surgical education must be explored.