Revista de Investigación en Tecnologías de la Información

Editorial for the special issue of the 13th International Conference on Software Engineering Research and Innovation

Juan Carlos Guzmán Preciado — 2025-10-26

Software Engineering focuses on analyzing, designing, implementing, testing, and maintaining software developments. It has become a strategic field since software is now present in almost every aspect of human life. We live in an era of constant technological advances, where software drives the digital transformation of society, industry, and academia. Given the importance of Software Engineering, this special issue presents the research articles in Spanish selected for the 13th International Conference on Software Engineering Research and Innovation (CONISOFT’25). The contributions reflect emerging trends, current challenges, and innovative solutions aimed at strengthening the scientific community, fostering collaboration between academia and industry, and advancing knowledge in the field of Software Engineering both nationally and internationally.

Initial results of a qualitative evaluation of the adherence of a strategy to improve the implementation and management of regression testing in small software development organizations

Edgar López Cruz — 2025-10-26

Regression tests running on a software product ensure that any modifications carried out do not generate new errors that negatively affect already implemented functionalities. However, defining a strategy for managing this type of testing is often difficult in small-sized software development organizations, as most face limitations in terms of financial resources, trained personnel, support tools, and experience. Therefore, this study introduces a strategy focused on these types of organizations based on the principles of Software Process Improvement and Knowledge Management. In this regard, as a first phase prior to implementing this strategy in real-life contexts, the design and implementation of a qualitative evaluation involving 135 software industry professionals is presented. These professionals determined the potential suitability of this strategy in the context of small organizations. The results obtained allowed for an evaluation of the strategy and established guidelines for designing a supporting computational tool to facilitate its empirical evaluation in these types of organizations.

Modeling computer systems by successive representations

Israel Sandoval Grajeda — 2025-10-26

The purpose of this text is to present a methodology to software development based on abstractions to build an informatic system based on a given model. This idea came to be as a result of certain challenges I faced while working as a software developer on a multidisciplinary project, where the main purpose is to analyze the behavior of a social phenomenon using an informatic tool that could mirror its behavior. This work reunites products of my eight years of experience as a collaborator in projects of software development, five of those in the aforementioned project, this includes my own definitions of concepts related to modeling and development process.

Systematic literature review: Intelligent tutoring systems

Mauricio Aburto Lara — 2025-10-26

Learning to program involves a steep learning curve, especially in the early stages, due to cognitive load, concept abstraction, and the inherent difficulties of programming languages. This study aims to identify recent technological approaches that integrate Intelligent Tutoring Systems (ITS) and Large Language Models (LLM) to support the programming teaching process. A systematic review was conducted following Kitchenham’s guidelines, consulting specialized databases, and applying inclusion and exclusion criteria in three phases to select relevant works in the field. The analyzed studies include proposals featuring conversational feedback, adaptive learning, and automated code analysis, showing improvements in concept comprehension, increased confidence, and higher task completion rates among beginner students. However, challenges were identified, such as the difficulty in maintaining context during prolonged interactions and the presence of erroneous responses or “hallucinations” in the models. It is concluded that multimodal integration, user-centered design, and optimized data management represent key areas to enhance the personalization and effectiveness of these systems in educational environments, fostering both skill development and continuous learning monitoring.

Property-based software testing automation through prompt engineering and artificial intelligence

Ricardo Rafael Quintero Meza — 2025-10-26

The increasing complexity of modern software development demands more efficient, comprehensive, and adaptive testing methodologies that ensure the reliability, robustness, and quality of applications, while also optimizing the costs and time associated with system maintenance and evolution. Although traditional testing approaches remain widely used, they present limitations in terms of coverage, scalability, and adaptability, especially when dealing with dynamic and constantly evolving systems. In this context, the integration of large language models (LLMs) through prompt engineering techniques emerges as a promising and innovative alternative for automating, enhancing, and expanding software testing processes. This work presents a tool that combines the generative capabilities of LLMs, accessible through specialized APIs, with the rigor of property-based testing (PBT). This synergy enables the automatic generation of test properties and the intelligent validation of code, facilitating early error detection and contributing to the development of more robust and reliable software from the early stages of the development lifecycle. Through prompt engineering, the tool guides the precise formulation of test properties and orchestrates the generation of diverse and relevant data. This approach aims to overcome the limitations of traditional methodologies by improving test coverage, reducing manual effort, and increasing scalability. The result is a more optimized verification process that promotes higher standards of software quality and reliability. This proposal represents a step forward in the intelligent automation of testing, integrating artificial intelligence with formal validation methodologies and opening new possibilities for its application in software engineering.

Challenges in urban environmental data visualization techniques from an adaptive perspective: a systematic review

Luis Alfredo Rojano-Ruiz — 2025-10-26

Data visualization faces the challenge of transforming into a dynamic and adaptable component to overcome its static limitations, responding to changing contexts and specific user needs. This systematic review aims to analyze how data visualization techniques have advanced toward more dynamic and flexible approaches. Tools, methods, and frameworks were identified that integrate the concept of adaptability as a response to the lack of flexibility of these techniques. As a result of this review, 26 papers were identified that address the incorporation of the concept of adaptability into interaction and customization based on the user's context, with the aim of optimizing communication through visual representations. Therefore, the need to transform how data is displayed is highlighted, moving from a static representation to a dynamic and adaptable tool, capable of responding to changing contexts and user needs.

Performance comparison of LLMs in the Generation of UML class diagrams using a RAG system

Lorena Martínez Sixto — 2025-10-26

This study presents a comparative evaluation of the performance of six LLMs (Gemini-2.5-Pro, Claude-Sonnet-4, GPT-4, DeepSeek-R1, Llama3.2, and Mistral) within a RAG system for generating UML class diagrams at the analysis level, based on user stories written in natural language. PlantUML code was used for the generation and evaluation of the diagrams, enabling a comparison between the generated diagrams and reference diagrams using the ROUGE-L metric, which focuses on average recall. The results showed that Gemini-2.5-Pro, Claude-Sonnet-4, and GPT-4 achieved better performance, with Claude-Sonnet-4 standing out by obtaining the highest average scores in most user stories. In contrast, DeepSeek-R1, Llama3.2, and Mistral presented difficulties, including the generation of invalid PlantUML code, which limited automated evaluation in some cases. The incorporation of the RAG system provided an initial foundation for exploring improvements in response quality, suggesting limitations in the quality and relevance of the retrieved context. Finally, opportunities for improvement were identified, such as prompt refinement and enhancement of the context used by the RAG system.

Improving effort estimation in software projects using oversampling and machine learning methods

Beatriz Bedolla Martínez — 2025-10-26

Effort estimation prediction determines the time it will take to develop a software program or the resources required to complete it within the established timeframe. A current alternative for predicting estimates is to use machine learning methods. However, publicly available data sets generally contain few samples, so such methods cannot improve their effectiveness. Thus, it is necessary to increase the number of samples using oversampling methods. Therefore, this paper presents the use of ensemble methods with combinations of oversampling and undersampling to analyze the performance impact of the regressors used on small and medium-sized data sets. Moreover, their effectiveness in improving effort estimation in software projects using measures such as MMRE, MAE, RMSE, and Pred is also presented. The results obtained from MMRE and Pred, mainly show that the application of these strategies reduces prediction errors. Consequently, the use of an appropriate ensemble model, together with oversampling and undersampling strategies, improves effort prediction, especially on small data sets such as COCOMO, Maxwell, and Desharnais with highly unbalanced sample distributions.