Research Context: Digital government increasingly relies on large-scale data processing for public services. Ensuring citizens’ privacy while enabling data-driven value creation is a critical challenge for Information Systems (IS) in the public sector. Scientific and/or Practical Problem: Traditional safeguards (e.g., access control, anonymization) are insufficient against reidentification risks and inter-organizational data sharing demands. Public agencies lack actionable guidance to select and deploy Privacy-Enhancing Technologies (PETs) fitting legal, organizational, and technical constraints. Proposed Solution and/or Analysis: We synthesize categories and application patterns of PETs (Differential Privacy, Secure Multiparty Computation, Homomorphic Encryption, Federated Learning, Trusted Execution Environments, Synthetic Data) and analyze their suitability to government scenarios. We provide policy-to-PET mapping using recent Brazilian federal decrees. Related IS Theory: Grounded in Sociotechnical Systems (alignment of people, processes, and technologies), Privacy by Design as a strategy, and Information Governance for accountability, transparency, and risk management. Research Method: Concept-centric analysis of PETs and their governance implications; document analysis of 2025 executive decrees mentioning personal data; analytic generalization to derive PET selection rationales for public-sector IS. Summary of Results: We (i) clarify privacy vs personal data protection for design, (ii) categorize PETs by function and lifecycle stage (data in use, input/output privacy), (iii) derive PET recommendations for inter-agency collaboration, secure analytics, and transparency (e.g., MPC/HE for cross-entity processing; Differential Privacy for open statistics), and (iv) identify capability and governance gaps (skills, interoperability, stewardship). Contributions and Impact to IS area: We bridge PETs’ technical capabilities with IS governance needs in digital government, offering a rationale for PET selection under regulatory, organizational, and sociotechnical constraints. The study advances responsible innovation in IS, informs publicsector architectures, and aligns with Brazil’s GranDSI-BR (2016–2026) by addressing ethics, transparency, and societal impacts of intelligent IS.
Research Context: Public policies such as the 'Pé-de-Meia' program are crucial for reducing school dropout in Brazil, but their evaluation is complex due to vast amounts of data and inherent selection bias. Scientific and/or Practical Problem: There is a need for an approach that moves beyond simple causal inference to identify the contextual factors that enhance the program's success, providing actionable intelligence for public managers. Proposed Solution and/or Analysis: We developed an intelligent information system that integrates public data and uses a Machine Learning model (LightGBM) to identify the main predictors of success in student retention, complemented by a quasi-experimental analysis (PSM) to estimate impact. Related IS Theory: This work is grounded in the principles of Decision Support Systems (DSS), applying computational intelligence to transform raw governmental data into strategic insights for policy management. Research Method: The methodology involved a four-stage data processing pipeline: data collection (Censo Escolar, Portal da Transparência), feature engineering (creating a proxy for school enrollment dynamics), predictive modeling with LightGBM to rank success factors, and impact estimation using Propensity Score Matching. Summary of Results: The predictive analysis identified that, beyond the natural inertia of previous performance, the number of teachers and program intensity are the most significant actionable predictors of success. To ensure robustness, these factors were cross-validated using a Random Forest model. The quasi-experimental analysis, refined with strict common support and caliper matching, estimated a positive and statistically significant impact (ATT = +0.19 p.p., p-value < 0.001). Contributions and Impact to IS area: This study contributes a novel, rigorously validated method for public policy evaluation within the IS field. It demonstrates how a predictive system can provide immediate strategic insights, while advanced quasi-experimental techniques can detect causal effects even in early-stage implementations, offering a tangible artifact for evidence-based management.
Research Context: Social media expansion has increased digital interactions, including hate speech such as racism, sexism, and homophobia. This challenges society and platforms to develop strategies for identifying and moderating harmful content; Scientific and/or Practical Problem: Despite AI advances, automatic detection faces limitations due to linguistic nuances, cultural context, and implementation costs. Scientifically, the challenge is evaluating model effectiveness; practically, it is developing economical, reliable large-scale solutions; Proposed Solution and/or Analysis: We conducted a comparative analysis of LLMs (GPT-3.5-Turbo, GPT-4.0, DeepSeek-V3 and Gemini-2.0-Flash) in detecting offensive social media comments. Tests on raw and preprocessed data using standardized prompts measured precision, cost, and execution time; Related IS Theory: The study draws on Information Systems theories, emphasizing socio-technical, ethical, and cost–benefit aspects (Socio-technical Theory, Actor-Network Theory, Resource-Based View, Dynamic Capabilities); Research Method: Over 2,000 comments were analyzed by LLMs using precision, recall, F1-score, operational cost, and processing time metrics; Summary of Results: GPT-4.0 achieved the highest F1-score (94.19%) but at high cost (US$ 26.99). DeepSeek-V3 balanced performance and cost (F1-score 93.37%, US$ 0.66). Gemini-2.0-Flash was the cheapest (US$ 0.12) but showed inconsistent results; Contributions and Impact to IS area: This work offers a practical framework for selecting LLMs for hate-speech detection based on accuracy, cost, and performance. It advances IS research by evaluating state-of-the-art models in real scenarios and providing guidance for ethical and efficient content moderation.
Research Context: The foreign exchange (Forex) market is the largest and most liquid in the world, characterized by high volatility and continuous operation. In this environment, human decision-making is often compromised by emotional biases and the limited ability to process large amounts of information in real time. Scientific and/or Practical Problem: Traditional trading strategies have weaknesses, including delayed signals and a lack of robust risk metrics. Furthermore, many existing studies in the literature focus only on cumulative return, neglecting the risk-return trade-off and practical applicability. Proposed Solution and/or Analysis: This work presents Scalper Major, an automated trading system designed to operate consistently in the Forex market. Its modular architecture integrates technical and managerial heuristics, as well as strict risk and capital management mechanisms. Related IS Theory: The research is based on the principles of information systems, applied to automated decision-making, aligning technical indicators, computational heuristics, and financial metrics as reliable support tools for investors. Research Method: The system was implemented in MQL5 on the MetaTrader 5 platform. The evaluation was conducted through eight-year backtests on four major currency pairs, taking into account commissions, execution delays, and various market scenarios. Summary of Results: With an initial capital of $20,000.00, Scalper Major achieved significant results: a net profit of $751,533.23, a win rate of 82.54%, and a maximum drawdown of 9.43%. The Sharpe Ratio of 1.63 demonstrates superior risk-return efficiency compared to related studies. Contributions and Impact to IS area: The study, in addition to proposing an operational tool, presents a methodological advancement by creating a novel compilation of the main evaluation indicators for automated trading systems.
Research Context: Observatories are information systems that support project monitoring, providing transparency and decision-making support through data collection, processing, storage, and dissemination. Despite their potential, the practical application of conceptual models for project observatories remains limited. Scientific and/or Practical Problem: The third version of the Model for Project Observatories (MPO) had not been assessed in real-world contexts, raising questions about its applicability, completeness, and usability. Proposed Solution and/or Analysis: This study evaluates the MPO through two multiple-case studies, examining the application of its concepts, identifying gaps, and suggesting improvements, with emphasis on usability aspects. Related IS Theory: The MPO builds on a sociotechnical perspective and relates to theories of conceptual modeling and project management support systems, highlighting transparency and governance in information systems. Research Method: A multiple-case study was conducted, following a qualitative approach. Data were collected through interviews, participant observation, document analysis, and source code review, allowing the assessment and refinement of the model. Summary of Results: The results show that most MPO concepts were successfully applied. Some concepts were only partially incorporated, and new elements were suggested to enhance usability. Cross-case analysis provided insights into strengths and limitations of the model. Contributions and Impact to IS area: This study empirically validates and refines the MPO, reinforcing its role as a conceptual foundation for project observatories. It contributes to the IS field by demonstrating the use of observatories as tools for transparency and project management, and by offering guidelines for the effective implementation of conceptual models.
Research Context: Software ecosystems (SECO) differ from other software development contexts and increase the complexity of risk management. Scientific and/or Practical Problem: It remains unclear the way risk management is conducted in SECO, which risks are specific to this context, and which solutions have been proposed. Proposed Analysis: This study investigates risk management in SECO by identifying SECO-specific risks, mapping solutions and their evaluations, and highlighting existing challenges. Related IS Theory. Socio-technical Theory provides a perspective by emphasizing the interdependence of technical and social factors in risk management in SECO. Research Method: We conducted a systematic mapping study following established guidelines, including study selection, data extraction, and qualitative coding. Summary of Results: We identified 24 SECO-specific risks, 21 solutions, of which 15 were evaluated, and 14 challenges related to risk management in SECO. Contributions and Impact to IS area: This study consolidates knowledge on risk management in SECO and provides insights to support researchers and practitioners in designing more effective mitigation strategies.
Research Context: Software testing in healthcare systems is critical, as failures can compromise diagnoses, professional trust and patient safety. In severe conditions, such as oral cancer, the absence of robust validation strategies may lead to delays or clinical errors with serious consequences. Scientific and/or Practical Problem: Despite advances in AI-based healthcare platforms, gaps remain in the adoption of integrated testing strategies that ensure quality, usability, and security in critical systems. These shortcomings can hinder the reliability of AI-driven tools in clinical environments. Proposed Solution and/or Analysis: This study reports the testing practices applied in the Sobre-Vidas Platform-“Oral Cancer”, designed to support early diagnosis through AI. A multidimensional testing strategy was implemented, encompassing back-end (unit, integration, performance, and security) and front-end testing (usability, responsiveness, accessibility and functional), supported by tools including Python, JUnit, SonarQube, Locust, and Cypress. Related IS Theory: The work is grounded in software quality models, particularly ISO/IEC 25010, and in socio-technical system theory, which emphasizes the interplay between technical reliability and user acceptance within digital health environments. Research Method: A qualitative case study approach was adopted, combining practices such as test driven development (TDD) and automated API testing with validation in real world healthcare contexts. Trials were conducted with physicians, dentists, and health secretaries in two municipalities in the state of Goiás, with feedback collected through observation, interviews, and usability sessions. Summary of Results: The need for adjustments in workflows, terminology, and interface design was revealed, highlighting that continuous and integrated testing processes are essential for ensuring clinical effectiveness, safety, and system adoption. Contributions and Impact to IS Area: This work demonstrates how integrated testing strategies, aligned with established quality frameworks, can enhance trust and adoption of AI-driven healthcare platforms. It broadens the discussion on software testing as a socio-technical enabler of safety and innovation in digital health.
Research Context: User satisfaction is recognized as a key success criterion in Information Systems (IS) projects. In the IS project context, continuously monitoring this criterion throughout the project lifecycle can reduce training and rework costs, enhance user experience, and prevent client loss. Scientific and/or Practical Problem: Despite its relevance, there is a lack of methods in the IS project management literature to evaluate user satisfaction during the development cycle. Existing approaches focus mainly on end-stage evaluations, limiting timely adjustments and evidence-based decision-making during project execution. Proposed Solution and/or Analysis: This paper proposes a framework for assessing user satisfaction in IS projects through three stages: expectation elicitation, expert inspection, and post-use verification. All stages are supported by questionnaires derived from usability heuristics to guide structured evaluation and comparison. Related IS Theory: The study is grounded in the Information Systems Success Model, which identifies user satisfaction as an essential dimension of project success, and integrates principles from Human–Computer Interaction (HCI). Research Method: An exploratory review of relevant literature was conducted, followed by the design of the proposed framework and its validation through three instrumental case studies across different organizational domains. Summary of Results: The results indicate that the proposed framework enables the evaluation of user satisfaction during system development, allowing early identification of usability issues and improvement opportunities. Contributions and Impact to the IS Area: This work contributes to IS project success management by operationalizing user satisfaction as a continuous and measurable criterion, ensuring its systematic verification throughout the project lifecycle and providing managers with a practical tool for evidence-based decision support.
Research Context: Educational repositories gather diverse Open Educational Resources (OER), yet sparse metadata and inconsistent terminology reduce findability. Large Language Models (LLMs) with retrieval-augmented generation (RAG) can bridge vocabulary gaps by capturing semantic similarity, thereby improving recall and user experience. Scientific and/or Practical Problem: The national OER repository (ProEdu) depends on a solely lexical engine. This dependence creates difficulties in handling synonyms, paraphrases, and domain shifts, resulting in suboptimal recall and inconsistent rankings. Proposed Solution and/or Analysis: We develop a prototype of an Educational Conversational Agent (ECA) that integrates retrieval and response generation. Three pipelines are evaluated: ProEdu, which employs a lexical approach; a field-weighted Elasticsearch (ES); and a semantic RAG system utilizing Sentence-Transformers (all-MiniLM-L6-v2) embeddings with a FAISS (Facebook AI Similarity Search) index and Llama for text generation, plus a lightweight reranking mechanism. Related IS Theory: We assert that AI-enhanced repositories diminish search obstacles and assist educators in effectively identifying suitable materials. Furthermore, the conversational interface alleviates the cognitive load by providing verified sources within context. Research Method: A comparative assessment involved 22 interdisciplinary prompts in ten domains. For each prompt, we established gold-standard datasets, formulated standardized queries, and calculated precision, recall, and F1-score. Summary of Results: The Llama/FAISS pipeline achieves the best coverage-relevance balance driven by high recall. ES attains a similar F1 through higher precision but lower recall. ProEdu performs poorly in F1. Error analysis shows semantic retrieval excels in cross-vocabulary matches and multi-facet intents. Contributions and Impact to IS area: We deliver a replicable benchmark for large-scale OER search (prompts, metrics, code) and a pragmatic architecture combining semantic RAG and ES to balance recall and precision on cost-efficient infrastructure. Prompt templates and evaluation scripts support adoption.
Research Context:Automatic text summarization remains a subject of considerable relevance across multiple domains. In particular, extractive multi-document generic summarization has garnered increased attention due to its capacity to mitigate information overload in a wide range of applications. Scientific and/or Practical Problem: The volume of unstructured text data produced on the internet has grown exponentially in recent years, driven by advances in information and communication technologies (ICTs). This massive generation of data makes it difficult for users to find relevant information. Proposed Solution and/or Analysis: This study introduces, implements, and applies the memetic algorithm known as Holistic Text Summarization with the Shuffled Frog-Leaping Algorithm (HSSFLA) to address the generic extractive multi-document multi-language text summarization problem using combinatorial optimization techniques. Related IS Theory: This research integrates swarm intelligence, memetic algorithms and combinatorial optimization. Research Method: An in vitro experiment was conducted to quantitatively compare the summary quality between the proposed method and similar methods in the literature. Summary of Results: Experiments were carried out on the DUC2001/2002 benchmark datasets, and performance was evaluated using the Recall-Oriented Understudy for Gisting Evaluation (ROUGE) metric. The results demonstrate that the proposed approach yielded an average improvement of 25.12% in ROUGE-1 and 34.91% in ROUGE-2 on the DUC 2001 dataset. On the DUC2002 dataset, the method achieved average gains of 35.42% in ROUGE-1 and 36.08% in ROUGE-2. Contributions and Impact to IS area: HSSFLA, a memetic algorithm based on swarm intelligence, was developed to solve this problem for the first time. It creates holistic summaries, in which it evaluates the quality of the summary as a whole, rather than focusing exhaustively on finding the best individual sentences. HSSFLA outperforms the results of the scientific literature in DUC2001 and DUC2002.
1) Research Context: The integration of advanced Machine Learning (ML) models into Intelligent Information Systems (IS) has created highly accurate but opaque "black-box" systems, especially in sensitive domains like credit scoring. 2) Scientific and/or Practical Problem: This opacity undermines user trust, can perpetuate algorithmic bias, and challenges regulatory compliance (e.g., LGPD, GDPR). This creates a critical gap between AI’s technical power and the socio-technical need for accountability in IS. 3) Proposed Solution and/or Analysis: We propose and validate a two-layer framework that uses Large Language Models (LLMs) to translate technical outputs from Explainable AI (XAI) methods, like SHAP and LIME, into actionable, natural language narratives for non-expert users. 4) Related IS Theory: Grounded in Decision Support Systems (DSS) theory, this work extends the classical DSS goal. It enhances decision quality not just via predictive accuracy, but by improving the transparency, trustworthiness, and interpretability of the system’s reasoning for stakeholders. 5) Research Method: We conducted an applied, experimental study on a public retail credit dataset. The methodology involved data preprocessing, XGBoost predictive modeling, quantitative evaluation of explanation fidelity with the MEMC metric, and developing a functional web prototype. 6) Summary of Results: The framework effectively identified key credit denial factors with high fidelity, validated by the MEMC metric. The LLM-synthesis layer successfully transformed complex XAI data into clear, understandable, and practical explanations, enhancing the system’s clarity and actionability. 7) Contributions and Impact to IS area: This study contributes a validated framework for building more ethical, transparent, and socially inclusive intelligent systems. Its impact lies in bridging the gap between advanced AI and human-centric requirements, enabling responsible AI adoption and strengthening human-AI collaboration in decision-making.
Research Context: The growing use of Machine Learning (ML) techniques in the development of predictive models to support decision-making and the development of information systems, while fostering advancements, has introduced significant ethical challenges, notably the emergence of biases that may lead to unfair decisions. Scientific and/or Practical Problem: The main scientific and practical challenge addressed is the identification and mitigation of algorithmic biases that may reproduce or exacerbate discrimination against minority groups, with a specific focus on the sensitive attribute gender. Proposed Solution and/or Analysis: This study presents an experimental evaluation of different bias mitigation strategies, including the use of the EqOddsPostprocessing and Reweighing methods from AI Fairness 360 toolkit, the application of weights, and the randomization of sensitive attribute values. Fairness performance was assessed using the Equal Opportunity and Demographic Parity metrics. Related IS Theory: This research is grounded in the theory of Algorithmic Fairness, aimed at ensuring impartiality, and in the concept of SocioTechnical Bias, which acknowledges that socially embedded prejudices are reflected in the outcomes produced by ML algorithms. Research Method: An experimental evaluation was conducted on a binary classification problem using the Portuguese SATDAP dataset. The baseline model was built with the Decision Tree algorithm, chosen for interpretability. Methodology comprised five experimental scenarios designed to test mitigation strategies and assess fairness through cross-validation. Summary of Results: Findings showed that Reweighing method fostered fairer predictions according to fairness metrics, in addition to yielding a slight but notable improvement in performance metrics such as accuracy, precision, recall, and f1-score. Contributions and Impact to IS area: This research reinforces the importance of integrating ethical guidelines and bias mitigation methodologies in developing ML systems, contributing to the construction of solutions fostering predictive fairness and countering discrimination without negatively impacting predictive performance.
Research Context: This article investigates software vulnerabilities and weaknesses across 273 machine learning (ML) library repositories. Scientific and/or Practical Problem: The scientific problem lies in comprehensively mapping security issues within the rapidly expanding landscape of ML libraries. Practically, this research addresses the need for a scalable methodology to effectively track, collect, and correlate existing vulnerabilities and weaknesses in these libraries. Proposed Solution and/or Analysis: We propose a scalable methodology designed to track, collect, and correlate security issues by leveraging CVEs (Common Vulnerabilities and Exposures) and CWEs (Common Weakness Enumeration). Related IS Theory: This research draws upon Complexity Theory, recognizing the intricate interconnectedness of ML ecosystems, where a single change or vulnerability can trigger cascading and unpredictable effects throughout the entire system. Research Method: Our research employed an empirical software analysis approach, involving the mining of 273 machine learning library repositories. We utilized CodeQL and Dependabot as part of the methodology to systematically track, collect, and conduct an in-depth analysis of the interconnections between CWEs and CVEs related to identified vulnerabilities and weaknesses. Summary of Results: Our findings revealed various security flaws, including validation failures, access control issues, memory management errors, development flaws, path-traversal vulnerabilities, and cryptographic weaknesses. Notably, these included CWEs from the Top 25 Most Dangerous Software Weaknesses. These results underscore the critical need for proactive measures to enhance security and reliability within ML systems. Contributions and Impact to IS area: Contributions include an automated methodology for characterizing vulnerabilities and weaknesses in ML libraries; an analysis of vulnerabilities and weaknesses in 273 ML libraries, and a dataset of CVEs/CWEs interconnections for ML.
Research Context: Large Language Models (LLMs) have emerged as promising resources to support software development. Beyond code generation, these tools can serve as technical and educational support, assisting developers in learning and applying concepts in complex low-code development platforms (LCDP). Scientific and/or Practical Problem: While LCDPs lower entry barriers, organizations still struggle to adopt them. As a first step, we investigate whether LLMs can effectively assist LCDP’s newcomers in learning, producing and deploying quality enterprise applications. Proposed Solution and/or Analysis: This paper presents an exploratory study on the use of ChatGPT to support the development of a Salesforce application. We focus on a zero-shot learning approach to resemble the profile of newcomers. The study aims to identify the suitability of LLMs to assist beginner low-code developer. Related IS Theory: ChatGPT is seen as part of a socio-technical network that transforms low-code development. Information systems are complex and require interoperability; in this context, generative AI expands accessibility and innovation. Research Method: Development of a low-code application through prompts, using ChatGPT as support. Evaluation of the application based on quantitative and qualitative metrics. Summary of Results: Simple requirements were often solved with a single prompt, whereas the most complex required iterations and significant debugging, revealing the model’s limits. Expert reviewers found the AI-generated solutions correct and aligned with Salesforce best practices, but noted that the code lacked structural best practices. Contributions and Impact to IS area: Contributes to SI by demonstrating that ChatGPT is a co-creative, educational partner that can boost development productivity and democratize development by lowering technical barriers. However, human expertise remains critical. This work demonstrates how strategic AI integration can enhance system delivery while supporting developer learning.
Research Context: Medical Information Systems (MIS) have evolved significantly to achieve efficiency and scalability in processing large amounts of multimodal data. Computer-Aided Diagnosis (CAD) models offer a second opinion to help physicians compose a more precise diagnosis. Effective CAD models need to be built on diverse, multicentric and, in specific cases, multimodal databases. Scientific and/or Practical Problem: The lack of standardization and complex interoperability among different MIS are major challenges that hinder the exploration of the maximum potential of digital data in healthcare applications, including the development of CAD models. Proposed Solution and/or Analysis: A generic, flexible and reusable relational data model was applied, developed specifically to support the training of CAD models from multimodal data. Through a case study using Cardiac Magnetic Resonance exams, a prototype with interactive Information Visualization (IV) functionalities was developed and qualitatively evaluated. Related IS Theory: The evaluation was inspired by the Fit-Viability Theory, as the prototype was evaluated primarily with regard to its potential for effective integration into real-world clinical workflows. Research Method: The research included: literature review; modeling; prototyping; development; qualitative evaluation through interviews with health professionals; and analysis and summarization of results. Summary of Results: Evaluated IV tools demonstrated strong potential to enhance clinical decision-making for the interviewed health professionals. From a Fit-Viability perspective, the system showed a promising fit with the users’ analytical needs and clinical workflows, while also indicating viability in terms of technical feasibility and resource compatibility within the healthcare setting. Contributions and Impact to IS area: Results showed that the strategic combination of a flexible data model and customizable IV functionalities can provide valuable research tools and diagnostic aids to physicians. These insights offer practical guidance for enhancing MIS and CAD capabilities, particularly in supporting the effective analysis of medical databases.
Research Context: The development of automated Information Systems (IS) capable of segmenting the left ventricle (LV) in cardiac magnetic resonance imaging (MRI) and estimating clinically relevant biomarkers is fundamental to support diagnostic decision-making. Scientific and/or Practical Problem: Many automated IS based on deep learning (DL) are not fully reliable and lack dedicated modules for error detection, which makes them dependent on constant manual inspection and correction. Proposed Solution and/or Analysis: We propose a post-processing method for IS that automatically detects and corrects segmentation errors in LV cardiac MRI produced by DL systems. Detection is performed by combining metrics computed between consecutive time frames of MRI to identify inconsistent segmentations in the temporal dimension; the correction step reconstructs the problematic frame by interpolating nearby segmentations. Related IS Theory: This work is grounded in the perspectives of Information Processing Theory. Research Method: The method was validated on LV segmentations containing both artificially generated and real DL errors, using the LVQuan19 dataset with reference segmentations for all time instants. Summary of Results: The method achieved detection performance with F1-score values up to 0.99 on real data, particularly for severe errors. Regarding the correction step, the selected strategy effectively improved segmentation consistency, achieving Dice coefficient values close to 0.95, indicating excellent agreement with reference segmentations. Contributions and Impact to IS area: This study contributes to the Information Systems field by introducing a method that improves the reliability of automated computer-aided systems for diagnosis. Although validated in the context of LV segmentation, the proposed approach can be applied to other domains where sequential data consistency is critical.
Context: Software development involves several steps and activities. One of these activities is task allocation. This activity is related to project management and is decisive for the success or failure of the project since it involves risks related to time and, consequently, costs. Practical Problem: Inefficient task allocation, combined with team member rotation and knowledge concentration among a few members, can negatively impact deadlines, costs, and even lead to project discontinuation. Proposed Solution: All knowledge related to the project must be distributed equally among all team members to mitigate the negative impacts of possible departures of team members. This work proposes a web solution for task allocation that considers the knowledge level of the team members on the project repository. IS Theory: The approach builds on theories of project management, knowledge distribution, and collaborative software development, emphasizing metrics for team knowledge assessment. Research Method: A web application implementing the proposed approach was developed and evaluated in the context of software development teams. Summary of Results: The solution enables balanced knowledge distribution among team members, mitigating risks associated with departures and knowledge concentration, and improves task allocation efficiency. Contributions: The proposed approach provides a practical tool for software project management, enhancing knowledge management and reducing project risks, contributing to both the research and practice of Information Systems.
Research Context: The emergence of Large Language Models (LLMs) has been transforming the development of information systems, directly impacting activities across the system life cycle. Understanding how these technologies affect work practices is essential for advancing the Information Systems (IS) field and for guiding their adoption by organizations. Scientific and/or Practical Problem: Despite the growing use of LLMs in information systems development, there is still limited knowledge on how professionals across different roles, levels of experience, and work modalities perceive the impact of generative AI on daily activities, decision-making, and team performance. Proposed Solution and/or Analysis: This study investigates the perceptions and experiences of Brazilian software developers regarding the integration of Generative AI (LLMs) into their professional practice. The analysis aims to identify which activities are most impacted by AI, the most frequently adopted tools, and the level of trust professionals place in AI-generated outputs. Related IS Theory: The study aligns with the Systemic and Socially Aware Perspective for Information Systems challenge of GranDSI-BR, by examining how emerging technologies reshape sociotechnical systems and the dynamics of professional work. Research Method: An online survey was conducted with 94 professionals in the field of information systems development, covering diverse profiles. Data were collected through customized Likert scales, multiple-choice checkboxes, and closed-ended questions. Summary of Results: Findings highlight that, although the adoption of LLMs significantly increases developers’ productivity, there is still a moderate level of trust in the outputs and the absence of consolidated governance policies. Contributions and Impact to IS area: The results provide both theoretical and practical insights into the adoption of generative AI in information systems development, offering valuable guidance for researchers, managers, and educators seeking to integrate these technologies into IS safely and ethically.
Research Context: The Brazilian Unified Health System faces overload due to high demand for emergency services. Primary Health Care (PHC) plays a central role in mitigating this issue but suffers from the uneven distribution of human resources. Scientific and/or Practical Problem: The lack of optimized strategies for professional allocation generates imbalances among PHC units, resulting in service deficits in some regions and underutilization in others. Proposed Solution and/or Analysis: This study proposes a decision support system that uses open health data and mathematical optimization models to reallocate professionals across primary healthcare units. The system is formulated as an Integer Linear Programming (ILP) problem, aiming to maximize the minimum Service Capacity Margin (MCA), which reflects the balance between supply and demand. Related IS Theory: The research draws on Information Systems theories focused on decision support, such as Decision Support Systems (DSS) Theory, and mathematical modeling approaches in organizational contexts. Research Method: An ILP model (MAX-MCA) was developed to allocate professionals by category in Primary Health Care units. Demand was estimated from sociodemographic data, distributed using Voronoi diagrams, and adjusted based on actual service production from public records. Supply was calculated from the weekly working hours of professionals and average productivity standards defined by the respective professional councils. Summary of Results: Experiments were conducted in the municipalities of Ribeirão Preto and Guarulhos. In both scenarios, the model eliminated deficits and increased MCA across all analyzed categories, demonstrating more balanced and efficient utilization of human resources. Contributions and Impact to IS area: This work illustrates how optimization models integrated with Information Systems can support decision-making, promoting greater equity and efficiency in resource allocation. The study contributes to advancing operational research applications in health management.
Research Context: Collaborative System of Information Systems (SoIS) involve multiple independent organizations that must interact dynamically, without centralized control, to achieve both collective and individual objectives. Organizational interoperability is crucial in this context but still lacks adequate solutions when participants frequently join or leave the collaboration. Scientific and/or Practical Problem: Current BPM and SOA approaches were designed for static intra or inter-organizational contexts and do not handle the dynamic nature of SoIS. This limitation creates difficulties in maintaining process continuity when participants enter or exit, compromising global objectives. Proposed Solution and/or Analysis: This research proposes a solution for organizational interoperability in collaborative SoIS comprising a metamodel and a semi-automated simulation environment. The metamodel specifies key concepts, properties, and relationships for modeling collaborative SoIS, while the simulation environment enables the specification and enactment of model instances according to the metamodel. Related IS Theory: Research on organizational interoperability is associated with organizational information processing theory and the complexity of information systems. Research Method: The Design Science Research (DSR) approach was adopted, following the steps of investigation, design, validation, implementation and evaluation. Summary of Results: The metamodel was applied in an e-commerce scenario, showing it to be consistent, expressive, and able to represent domain concepts through models adhering to its rules and structure. The simulation environment executed these models, enabling analysis of participant entry and exit dynamics and demonstrating process continuity despite organizational changes. Contributions and Impact to IS area: The approach provide both a conceptual foundation and a means for analyzing and improving collaboration in dynamic interorganizational contexts. The impact lies in enabling more resilient and adaptive SoIS, capable of maintaining continuity and interoperability even under frequent organizational changes.
