Artículos (Lenguajes y Sistemas Informáticos)

URI permanente para esta colecciónhttps://hdl.handle.net/11441/11392

Examinar

Mostrando 1 - 20 de 947

Code to learn: Where does it belong in the K-12 curriculum?
(Informing Science Institute, 2016) Moreno León, Jesús; Robles, Gregorio; Román González, Marcos; Lenguajes y Sistemas Informáticos; TIC276: Diverso Lab - International Computing
The introduction of computer programming in K-12 has become mainstream in the last years, as countries around the world are making coding part of their curriculum. Nevertheless, there is a lack of empirical studies that investigate how learning to program at an early age affects other school subjects. In this regard, this paper compares three quasi-experimental research designs conducted in three different schools (n=129 students from 2nd and 6th grade), in order to assess the impact of introducing programming with Scratch at different stages and in several subjects. While both 6th grade experimental groups working with coding activities showed a statistically significant improvement in terms of academic performance, this was not the case in the 2nd grade classroom. Notable disparity was also found regarding the subject in which the programming activities were included, as in social studies the effect size was double that in mathematics.
A Survey of Quantum Machine Learning: Foundations, Algorithms, Frameworks, Data and Applications
(ACM Digital Library, 2025) Rodríguez Díaz, Francesc; Gutiérrez Avilés, David; Troncoso Lora, Alicia; Martínez Álvarez, Francisco; Lenguajes y Sistemas Informáticos; Ministerio de Ciencia, Innovación y Universidades (MICIU). España; TIC254: Data Science & Big Data Lab
Quantum machine learning combines quantum computing with machine learning to solve complex computational problems more efficiently than classical approaches. This survey provides an introduction to the foundations, algorithms, frameworks, data and applications of quantum machine learning, serving as a resource for researchers and practitioners. We begin by reviewing existing surveys to identify gaps that this work addresses, followed by a detailed discussion of the foundational principles of quantum mechanics and machine learning essential for quantum machine learning. Key algorithms are examined, highlighting their mechanisms, advantages, and applications across various domains. Current frameworks and platforms for implementing quantum machine learning algorithms are explored, emphasizing their unique features and suitability for different contexts. Existing quantum datasets for practical usage are also reported and commented on. This survey also reviews over 135 articles, categorized into theoretical and practical contributions, to identify key advances, limitations, and application areas within quantum machine learning. Critical challenges such as hardware limitations, error rates, and scalability are analyzed to detect the obstacles that must be addressed for practical deployment. By synthesizing these elements into a structured overview, this survey aims at serving as both an introduction and a guide for advancing research and development in this disruptive field.
Research hypothesis generation over scientific knowledge graphs
(Elsevier, 2025) Borrego Díaz, Agustín; Dessì, Danilo; Ayala Hernández, Daniel; Hernández Salmerón, Inmaculada Concepción; Osborne, Francesco; Recupero, Diego Reforgiato; Buscaldi, Davide; Ruiz Cortés, David; Motta, Enrico; Lenguajes y Sistemas Informáticos; Ministry of University and Research (MUR)
Generating research hypotheses is a crucial step in scientific investigation that involves the creation of precise, verifiable, and logically valid statements that can be empirically examined. Therefore, many efforts have been made to automate or assist this process through the use of various Artificial Intelligence solutions. However, most existing methods are tailored to very specific domains, particularly within the biomedical field. There have been recent attempts to formalize hypothesis generation as a link prediction task over knowledge graphs. This solution is potentially domain-independent and applicable across diverse disciplines. Nevertheless, current approaches for link prediction, which typically rely on embedding models or path-based methods, have shown limited success in accurately predicting new hypotheses. To address these limitations, this paper introduces ResearchLink, an innovative and domain-independent methodology for hypothesis generation over knowledge graphs. ResearchLink combines path-based features and knowledge graph embeddings with text embeddings, capturing the semantic context of entities within a given corpus, and integrates additional information from bibliometric databases to improve research collaboration predictions. To conduct a rigorous evaluation of ResearchLink, we constructed CSKG-600, a new dataset for hypothesis generation, consisting of 600 statements that were manually labeled by domain experts. ResearchLink achieved outstanding performance (78.7% P@20), significantly outperforming alternative approaches such as TransH (71.8%), TransD (71.7%), and RotatE (70.7%).
A systematic review of capability and maturity innovation assessment models: Opportunities and challenges
(Elsevier, 2023) Giménez Medina, Manuel; González Enríquez, José; Domínguez Mayo, Francisco José; Lenguajes y Sistemas Informáticos; Ministerio de Ciencia e Innovación (MICIN). España; TIC276: Diverso Lab - International Computing
Public funding, being the primary source for innovation, imposes restrictions caused by a lack of trust between the roles of public funders and organisations in the innovation process. Capability and maturity innovation assessment models can improve the process by combining both roles to create an agile and trusting environment. This paper aims to provide a current description of the state-of-the-art on capability and maturity innovation assessment models in the context of Information and Communication Technologies. To this end, a Systematic Mapping Study was carried out considering high-quality published research from four relevant digital libraries since 2000. The 78 primary studies analysed show several gaps and challenges. In particular, a common ontology has not been achieved, and Innovation Management Systems are scarcely considered. Concepts such as open innovation have not been correctly applied to incorporate all Quadruple Helix stakeholders, especially the government and its role as a public funder. This implies that no studies explore a standard agile public–private maturity model based on capabilities since the public funders’ restrictions have not been considered. Furthermore, although some concepts of innovation capabilities have evolved, none of the studies analysed offer a comprehensive coverage of capabilities. As potential future lines of research, this paper proposes 11 challenges based on the 5 shortcomings found in the literature.
The innovation challenge in Spain: A Delphi study
(Elsevier, 2023) Giménez Medina, Manuel; González Enríquez, José; Olivero González, Miguel Ángel; Domínguez Mayo, Francisco José; Lenguajes y Sistemas Informáticos; Ministerio de Ciencia e Innovación (MICIN). España; TIC276: Diverso Lab - International Computing
Public funding for information and communication technology (ICT) innovation in Spain appears to be slow, bureaucratic, highly restrictive, and not agile. Therefore, the innovation process is negatively affected. These restrictions could be attributed to inadequate trust from public funders toward executors and ontological problems regarding the definitions of ICT innovation (i.e., the I+D+i formula), affecting all Quadruple Helix stakeholders. In this study, a Delphi study was proposed to reach a consensus among 81 experts (i.e., innovation managers, public funders, and consultants) to validate this hypothesis. The study included 41 statements and 59 questions organized into the following five objectives: (1) concept of innovation, (2) public funding and its restrictions, (3) theoretical model of innovation, (4) public funders’ trust and executors’ freedom, (5) assessment of capabilities and maturity for innovation. The experts discussed, evaluated, and reached a consensus, after two rounds, on 52 of the 59 questions. The results revealed wide dispersion of the proposed ICT innovation questions. They demonstrated that the innovation management ecosystem in Spain’s ICT context is immature and the I+D+i formula did not represent the innovation process. The study reached a consensus on requirements for an Agile Innovation Funding Framework (AIFF) oriented toward obtaining an improved competitive advantage for ICT products or services based on trust, transparency, inspection, and adaptation principles. The results revealed that a joint framework involving public funders and executors based on organizational capability and maturity positively affects the innovation process. The capabilities of the executors should be standardized and measured, and public funders must move from supervisors to mentors to acquire new capabilities. Furthermore, innovation regulation and the various types of calls for proposals should be analyzed globally to change their fiscal and controlling nature restricting innovation.
FallacyES-Political: A Multiclass Dataset of Fallacies in Spanish Political Debates
(SOC Española Procesamiento Lenguaje Natural-SEPLN, 2025) Cruz Mata, Fermín; Enríquez de Salamanca Ros, Fernando; Ortega Rodríguez, Francisco Javier; Troyano Jiménez, José Antonio; Lenguajes y Sistemas Informáticos; Ministerio de Ciencia, Innovación y Universidades (MICIU). España; TIC134: Sistemas Informáticos
Fallacies are pervasive in political discourse, shaping public opinion and influencing decision-making. Automatic detection and classification of fallacies is a challenging task, especially in non-English languages due to limited resources. In this study, we present FallacyES-Political, a novel dataset of fallacies extracted from 19 electoral debates held in Spain over three decades. The dataset comprises nearly 2,000 fallacies categorized into 16 types. To evaluate the dataset’s utility, we conducted a comprehensive benchmarking of state-of-the-art Large Language Models (LLMs) in zero-shot classification. The results highlight the complexity of fallacy classification and the limitations of current LLMs in understanding context-dependent argumentation. Furthermore, we demonstrate the advantages of fine-tuning a compact, domain-specific model over relying on general-purpose LLMs, achieving notable improvements in classification accuracy with a more sustainable approach.
NDT: a methodology to deal with the navigation aspect at the requirements phase
(2004) Escalona Cuaresma, María José; Reina Quintero, Antonia María; Torres Valderrama, Jesús; Mejías Risoto, Manuel; Lenguajes y Sistemas Informáticos; TIC258: Data-Centric Computing Research Hub
The proliferation of the internet has provoked the conception of new design methodologies to deal with the new concepts that have appeared in web applications and weren't treated in traditional methodologies. One of these concerns is navigation. It has been proved that traditional abstraction mechanisms (functions and objects) do not treat very well some concerns that are scattered all over the program code. In this paper, we think about navigation as an aspect that can be treated independently during the development of the system. We introduce a process to identify and define such an aspect during the requirements specification phase. This process is part of the NDT (Navigational Development Technique) development process, a new approach to develop web information systems.
Annotated Spanish general election debate transcriptions 1993-2023
(Nature Publishing Group, 2025) Cruz Mata, Fermín; Enríquez de Salamanca Ros, Fernando; Ortega Rodríguez, Francisco Javier; Troyano Jiménez, José Antonio; Lenguajes y Sistemas Informáticos; Ministerio de Ciencia e Innovación (MICIN). España; TIC134: Sistemas Informáticos
The automatic analysis of political discourse requires systematic resources, particularly for non-Anglophone electoral debates, which differ from parliamentary sessions. This paper describes the DebatES dataset, containing manually reviewed transcriptions of nearly all Spanish nationally televised general election debates since 1993. Transcripts are segmented into turns and thematic blocks and include participant metadata. The data is enriched with linguistic-stylistic metrics derived from NLP analysis and extensive annotations combining Large Language Models and manual validation. Annotations cover turn topics, emotions, relevant entity mentions, electoral proposals, and factual claims. To facilitate access and reuse, the dataset is distributed in standard formats (XML and CSV), accompanied by interactive reports for visual exploration. This dataset provides a valuable resource for researchers in linguistics, political science, and language technologies, opening new avenues for studying the evolution of discourse, persuasive strategies, and ideological polarisation within the Spanish political context.
Software bug report dataset from Eclipse projects
(Elsevier, 2025) López Durán, Noelia; Romero Organvidez, David; Cruz Mata, Fermín; Benavides Cuevas, David Felipe; Lenguajes y Sistemas Informáticos; Ministerio de Ciencia e Innovación (MICIN). España; TIC134: Sistemas Informáticos
In recent decades, the analysis of data from software projects including source control systems, defect tracking systems, and code review repositories has greatly improved our understanding of software development and its evolution. How ever, obtaining this information can be time-consuming, and the extracted data is not always well-maintained. This paper introduces an extensive dataset generated from Bugzilla repositories, focusing on key products from the Eclipse bug-tracking system. This dataset addresses the need for up-to- date data in existing repositories, preserving crucial historical information that may be lost due to the transition from Bugzilla to newer bug-tracking systems like Jira or GitHub Is- sues. Our dataset includes 301,378 bug reports along with all related information, organised into different folders that indicate the project in which the bug was filed. Additionally, we present a custom and lightweight Command Line Interface (CLI) tool designed to efficiently extract detailed information from Bugzilla repositories, automating data collection across various Bugzilla instances. The dataset and tool can be utilized for defect prediction, software maintenance, and evolutionary analysis. To the best of our knowledge, this is the largest, most complete, and up-to-date dataset of Eclipse bug reports available.
Beam energy forecasting using machine learning at the clear accelerator
(JACoW Publishing, 2025) Gilardi, Antonio; Aksoy, Avni; Bonnard, Lucile; Carranza García, Manuel; Corsini, Roberto R.; Farabollini, Wilfrid; Filippetto, Daniele; Franek, O; Gamba, Davide; Granados, Eduardo; Korysko, Pierre; Malyzhenkov, Alexander; Mazzoni, Stefano; Mostacci, Andrea; Petersson, A; Pollastro, A; Rieker, Vilde; Sjøbæk, K. N; Tangari, Giacomo; Wroe, Laurence M; Lenguajes y Sistemas Informáticos; PNRR MUR; Horizon Europe; TIC134: Sistemas Informáticos
Accurate and stable beam parameters are crucial for the success of particle accelerator-based experiments. Traditional methods for measuring beam parameters, however, often rely on invasive techniques that can disrupt experiments. This paper presents an initial step for obtaining a novel, non-invasive machine learning-based approach for predicting beam energy using parasitic measurements, enabling accurate real-time prediction without interference. The method employs a predictive model optimized for onestep-ahead forecasting and uses time-series data decomposition to handle complex beam energy dynamics, with recursive prediction strategies allowing it to anticipate future variations autonomously. Preliminary results from experiments at the CLEAR accelerator demonstrate the model’s ability to capture both slow trends and rapid energy shifts, and adapt to diverse experimental needs. These findings lay the foundation for future studies, emphasizing the potential of machine learning to measure beam energy and provide a real-time, non-destructive alternative to conventional methods. This approach promises significant advancements in accelerator-based applications, especially where destructive techniques are impractical.
CoSTI: Consistency models for (a faster) spatio-temporal imputation
(Elsevier, 2025-07) Solís García, Javier; Vega Márquez, Belén; Nepomuceno Chamorro, Juan Antonio; Nepomuceno Chamorro, Isabel de los Ángeles; Lenguajes y Sistemas Informáticos
Multivariate Time Series Imputation (MTSI) is crucial for many applications, such as healthcare monitoring and traffic management, where incomplete data can compromise decision-making. Existing state-of-the-art methods, like Denoising Diffusion Probabilistic Models (DDPMs), achieve high imputation accuracy; however, they suffer from significant computational costs and are notably time-consuming due to their iterative nature. In this work, we propose CoSTI, an innovative adaptation of Consistency Models (CMs) for the MTSI domain. CoSTI employs Consistency Training to achieve comparable imputation quality to DDPMs while drastically reducing inference times, making it more suitable for real-time applications. We evaluate CoSTI across multiple datasets and missing data scenarios, demonstrating up to a 98% reduction in imputation time with performance on par with diffusion-based models. This work bridges the gap between efficiency and accuracy in generative imputation tasks, providing a scalable solution for handling missing data in critical spatio-temporal systems.
Configuration Bugs Classification using LLMs and Encoders
(ACM Digital Library, 2025-08-31) López Durán, Noelia; Romero Organvidez, David; Cruz Mata, Fermín; Benavides Cuevas, David Felipe; Lenguajes y Sistemas Informáticos
Configuration-related bugs in software development are reported to be one of the main sources of problems in different domains and are especially relevant in software product lines. Classifying configuration bugs is difficult, since most of the bug tracking systems use descriptions written in natural language that have to be interpreted to categorize them. Some projects use labels to classify bugs, but others do not. In this paper, we address the challenge of automatically classifying bug reports as either configuration-related or non-configuration-related by leveraging the capabilities of large language models (LLMs). We conducted experiments involving zero-shot classification using general-purpose LLMs, as well as fine-tuning a smaller encoder-based model. These approaches were evaluated on two different datasets: first, a dataset consisting of 1,331 bug reports from a web-based software product line, manually labelled by us; and second, a refined version of an existing dataset from the literature, improved through additional data curation. The results are promising, with an F1 score of 0.81 using a combination of both datasets. Furthermore, we applied the fine-tuned encoder-based model to estimate the proportion of configuration-related bugs within a large, previously unseen dataset extracted from various highly configurable Eclipse projects, which contains over 300k bug reports. Our analysis yielded an estimated proportion of 36% for configuration-related bugs. To the best of our knowledge, this is the first work using LLMs to classify configuration bugs. Our results open a new perspective on configuration bugs classification and pave the way to investigate further in this direction.
A New Metric Based on Association Rules to Assess Feature-Attribution Explainability Techniques for Time Series Forecasting
(IEEE COMPUTER SOC., 2025-05) Troncoso-García, A. R.; Martínez Ballesteros, María del Mar; Martínez-Álvarez, F.; Troncoso, A.; Lenguajes y Sistemas Informáticos; Ministerio de Ciencia e Innovación (MICIN). España
This paper introduces a new, model-independent, metric, called RExQUAL, for quantifying the quality of explanations provided by attribution-based explainable artificial intelligence techniques and compare them. The underlying idea is based on feature attribution, using a subset of the ranking of the attributes highlighted by a model-agnostic explainable method in a forecasting task. Then, association rules are generated using these key attributes as input data. Novel metrics, including global support and confidence, are proposed to assess the joint quality of generated rules. Finally, the quality of the explanations is calculated based on a wise and comprehensive combination of the association rules global metrics. The proposed method integrates local explanations through attribution-based approaches for evaluation and feature selection with global explanations for the entire dataset. This paper rigorously evaluates the new metric by comparing three explainability techniques: the widely used SHAP and LIME, and the novel methodology RULEx. The experimental design includes predicting time series of different natures, including univariate and multivariate, through deep learning models. The results underscore the efficacy and versatility of the proposed methodology as a quantitative framework for evaluating and comparing explainable techniques.
Urban Pollution Impact Assessment in Six Lithuanian Cities With a Focus on Road Traffic Emissions - Integrated Framework for Environmental Health Studies
(Elsevier, 2025-10) Kecorius, Simonas; Madueño, Leizel; Birmili, Wolfram; Löndahl, Jakob; Plauškaitė, Kristina; Byčenkienė, Steigvilė; Lovrić, Mario; Petrić, Valentino; Carranza García, Manuel; Jiménez Navarro, Manuel Jesús; Martínez Ballesteros, María del Mar; Weiss, Magdalena; Schmid, Otmar; Cyrys, Josef; Peters, Annette; Kecorius, Gaudentas; Lenguajes y Sistemas Informáticos
An integrated framework is introduced and applied to assess the health impact of airborne pollution with greater physiological relevance, moving beyond conventional exposure metrics. Measured particle number size distribution data was integrated with a regional respiratory tract deposition fractions to estimate total and alveolar deposited particle surface area concentrations. Land use regression modeling, combined with randomized commuting patterns, enabled the evaluation of city-specific alveolar surface area deposition doses, providing new insight into localized average exposure and its implications for public health. The results showed that although the mean street-level air pollution in Lithuania is higher than in other European cities, the urban background levels are on the same level. We found that the total respiratory deposited surface area concentration is up to 18-fold higher due to coarse particles, which also determines alveolar deposited particle surface area dose. Our findings advocate for using integrated pollution assessments and region-specific policies rather than broad diesel vehicle-targeted bans. The proposed methodology is expected to enhance traditional exposure assessments by switching to lung deposited surface area, which can be further refined by incorporating daytime activity patterns, socio-economic status, and personal health conditions.
Road-traffic emissions of ultrafine particles and elemental black carbon in six Northern European cities
(Elsevier, 2025) Kecorius, Simonas; Madueño, Leizel; Plauškaitė, Kristina; Byčenkienė, Steigvilė; Lovrić, Mario; Petrić, Valentino; Carranza García, Manuel; Jiménez Navarro, Manuel Jesús; Martínez Ballesteros, María del Mar; Kecorius, Gaudentas; Lenguajes y Sistemas Informáticos
Urban air pollution from vehicular emissions remains a pressing public health concern, particularly in Eastern Europe, where data gaps hinder effective mitigation. This study, conducted in the summer of 2024, presents the first detailed analysis of ultrafine particle (UFP) and equivalent black carbon (eBC) emissions from road traffic across Lithuania’s six major cities: Vilnius, Kaunas, Klaipėda, Šiauliai, Panevėžys, and Alytus. We used a custom mobile laboratory to capture real-world emissions, revealing stark spatial disparities. Panevėžys and Vilnius topped eBC levels (10400 ng/m³ and 10200 ng/m³, respectively), driven by aging vehicle fleets and a diesel prevalence of 70 % in Panevėžys, which also recorded the highest UFP concentration (97800 particles/cm³). Emission factors, calculated using an adapted Operational Street Pollution Model (OSPM), identified Vilnius’ light-duty vehicles as leading in particle number emissions (8.90 × 10¹⁴ particles/(km·veh)), likely due to the prevalence of gasoline direct injection engines. At the same time, Panevėžys dominated eBC emissions (150 mg/(km·veh). Heavy-duty vehicles, including buses and trucks, exhibited emission factors up to five times higher than those of their light-duty counterparts, thereby amplifying their impact in urban areas. These findings illuminate emission dynamics in an understudied region, providing policymakers with precise and actionable insights for targeted interventions, such as fleet upgrades or the establishment of low-emission zones. By addressing a critical knowledge gap, this study empowers the scientific community and public health advocates to devise strategies that combat vehicle-related pollution, reduce exposure to harmful pollutants, and foster healthier urban environments across Eastern Europe and beyond.
IDE4ICDS: A Human-Centric and Model-Driven Proposal to Improve the Digitization of Clinical Practice Guideline
(Assoc. Computing Machinery, 2024-09-27) Parra-Calderón, Carlos Luis; García García, Julián Alberto; Ramos-Cueli, Juan Manuel; Alvarez-Romero, Celia; Román-Villarán, Esther; Escalona Cuaresma, María José; Lenguajes y Sistemas Informáticos
Clinical practice guidelines (CPGs) are a formalization of specific clinical knowledge that states the best evidence-based clinical practices for treating pathologies. However, CPGs are limited because they are usually expressed as text. This gives rise to a certain level of ambiguity, subjective interpretation of the actions to be performed, and variability in clinical practice by different health professionals facing the same circumstances. The inherent complexity of CPGs is also a challenge for software engineers designing, developing, and maintaining software systems and clinical decision support system to manage and digitize them. This challenge stems from the need to evolve CPGs and design software systems capable of allowing their evolution. This paper proposes a model-driven, human-centric and tool-supported framework (called IDE4ICDS) for improving digitisation of CPG in practical environments. This framework is designed from a human-centric perspective to be used by mixed teams of clinicians and software engineers. It was also validated with the type 2 diabetes mellitus CPG in the Andalusian Public Health System (Spain) involving 89 patients and obtaining a kappabased analysis. The recommendations were acceptable (0.61–0.80) with a total kappa index of 0.701, leading to the conclusion that the proposal provided appropriate recommendations for each patient.
COTriage: Applying a Model-Driven Proposal for Improving the Development of Health Information Systems with Chatbots
(Institute of Electrical and Electronics Engineers (IEEE), 2024-06-26) García García, Julián Alberto; Sánchez Gómez, Nicolás; Escalona Cuaresma, María José; Ruiz, Mercedes; Lenguajes y Sistemas Informáticos; Ministerio de Ciencia e Innovación (MICIN). España
Today, organizations require innovative and flexible solutions to digitize and automate their processes—particularly those processes designed to obtain information from users. Chatbots are one of the most widely used technological options for automating processes. This article aims to integrate chatbot technology with health information systems (HISs) to improve the execution of health-care processes. Specifically, it presents the COTriage framework, which proposes model-driven mechanisms for improving the design and development of process-oriented HISs with chatbot-based triage process integration. The proposal was also instantiated with the COVID-19 triage process and assisted reproduction treatment processes on iMedea (a real HIS). Finally, the proposal was discussed considering the acceptance of end users, as well as the degree of efficiency and effectiveness achieved by the software team who applies COTriage on our case study.
Pragmatic random sampling of Kconfig-based systems: A unified approach
(Elsevier, 2025-08) Fernandez-Amoros, David; Heradio, Ruben; Horcas Aguilera, José Miguel; Galindo Duarte, José Ángel; Benavides Cuevas, David Felipe; Fuentes, Lidia; Lenguajes y Sistemas Informáticos
The configuration space of some systems is so large that it cannot be computed. This is the case with the Linux Kernel, which provides more than 18,000 configurable options described across almost 1,700 files in the Kconfig language. As a result, many analyses of these systems rely on sampling their configuration space (e.g., debugging compilation errors, predicting configuration performance, finding the configuration that optimizes specific performance metrics, among others.). The Kernel and other Kconfig-based systems can be sampled pragmatically, using their built-in tool conf to get a sample directly from the Kconfig specification that is approximately random, or idealistically, generating a genuine random sample by first translating the Kconfig files into logic formulas, then using a logic engine to compute the probability that each option value has to appear in a configuration, and finally utilizing these probabilities to generate an authentically random sample. The pros of the idealistic approach are that it ensures the sample is representative of the population, but the cons are that it sets out many challenging problems that have not been solved yet (fundamentally, how to obtain a valid translation into Boolean that covers all the Kconfig language, and how to compute the option value probabilities for very large formulas). This paper introduces a new version of conf called randconfig , which incorporates a series of improvements that increase the randomness and correctness of pragmatic sampling and also help validate the Boolean translation required for the idealistic approach. randconfig has been tested on ten versions of the Linux Kernel and twenty additional Kconfig systems. Its compatibility significantly enhances the current landscape, where some systems use a customized conf variant that is maintained independently, while others do not support sampling at all. randconfig not only offers universal sampling for all Kconfig systems but also simplifies its evolutive maintenance as a single tool rather than an unorganized collection of conf variants.
A Conceptual Framework for Smart Governance Systems Implementation
(IGI Global, 2025) Muñoz-Hermoso, Salvador; Domínguez Mayo, Francisco José; Cerrillo-I-Martínez, Agustí; Benavides Cuevas, David Felipe; Lenguajes y Sistemas Informáticos; Ministerio de Ciencia e Innovación (MICIN). España; Universidad de Sevilla
Knowledge-based decision-making open to citizens holds little significance in governments. One reason is that so far, no reference frameworks are available to implement smart governance systems in the full public policy cycle, resulting in most of the existing tools not being knowledge-based. Thus, there is a risk that decisions are ineffective or misaligned with the different interests of civil society. Moreover, existing proposals do not cover most of the key features needed in smart governance and do not provide sufficient elements to facilitate its implementation. Based on the existing literature and tools, as well as on a survey of local government practitioners, the authors propose a conceptual framework for implementing smart governance systems, which manages both knowledge internal and external to the organization, and the one provided by stakeholders; thus, improving consensus and group decision-making. To this end, the framework considers available data and information technologies, and its components make it easier for institutions and information technology providers to develop solutions with a knowledge-based collaborative governance model.
Transformer and Adaptive Threshold Sliding Window for Improving Violence Detection in Videos
(MDPI, 2024-08-16) Rendón-Segador, F.J.; Álvarez García, Juan Antonio; Soria Morillo, Luis Miguel; Lenguajes y Sistemas Informáticos; Ministerio de Ciencia e Innovación (MICIN). España
This paper presents a comprehensive approach to detect violent events in videos by combining CrimeNet, a Vision Transformer (ViT) model with structured neural learning and adversarial regularization, with an adaptive threshold sliding window model based on the Transformer architecture. CrimeNet demonstrates exceptional performance on all datasets (XD-Violence, UCF-Crime, NTU-CCTV Fights, UBI-Fights, Real Life Violence Situations, MediEval, RWF-2000, Hockey Fights, Violent Flows, Surveillance Camera Fights, and Movies Fight), achieving high AUC ROC and AUC PR values (up to 99% and 100%, respectively). However, the generalization of CrimeNet to cross-dataset experiments posed some problems, resulting in a 20–30% decrease in performance, for instance, training in UCF-Crime and testing in XD-Violence resulted in 70.20% in AUC ROC. The sliding window model with adaptive thresholding effectively solves these problems by automatically adjusting the violence detection threshold, resulting in a substantial improvement in detection accuracy. By applying the sliding window model as post-processing to CrimeNet results, we were able to improve detection accuracy by 10% to 15% in cross-dataset experiments. Future lines of research include improving generalization, addressing data imbalance, exploring multimodal representations, testing in real-world applications, and extending the approach to complex human interactions.

Examinar

Envíos recientes