Sistema Regional de Información
en línea para Revistas Científicas de América Latina,
el Caribe, España y Portugal

Preface to the Iberoamerican Human-Computer Interaction Conference 2024 Special Issue

Gas consumption is a critical aspect of the pelletizing process, directly influencing operational costs and environmental impact. This study investigates the application of a multivariate time series forecasting pipeline for predicting gas consumption in pelletizing plants. The pipeline comprises: (i) data preprocessing, (ii) converting the dataset into a tabular format using a sliding window technique, (iii) applying feature selection methods, and (iv) employing machine learning tuned via AutoML. The methodology was tested on a dataset with 45 operational parameters collected over 90 days from an industrial plant, with predictions evaluated using Root Mean Squared Error (RMSE). In step (iii), twelve features were identified as the most relevant based on the Random Forest importance index. In the final stage, two AutoML approaches were employed: neural architecture search using AutoKeras and the DEAP (Distributed Evolutionary Algorithm) framework. The neural network architectures tested included MLP, RNN, LSTM, and Conv1D. The best performance was achieved by the DEAP framework combined with LSTM networks, which yielded an RMSE of 0.33. Although AutoML did not outperform the statistical model in terms of RMSE values, regarding training time, AutoML models were significantly more efficient than the statistical approach, optimizing computational resource usage and enabling faster model adjustments. These findings confirm the generalization capability of the pipeline, demonstrating its applicability across different industrial environments.

Context: Active learning provides learning contexts where students can acquire experience by doing contextualized and meaningful activities that involve high-order thinking. Despite active learning being proved to be more effective providing hands-on experience in software engineering situations, traditional learning activities such as lectures still rival the former. Objectives: Identify and define what active learning strategies are used in software engineering higher education (1) and what application experiences of such strategies have been conducted between 2018 and 2023 (2). Method: We conducted a Rapid Review to identify active learning strategies applied in Software Engineering in higher education. We conducted an empirical case study about the deployment and evaluation of active learning strategies in our Software Engineering Department. Results: 18 active learning strategies on software engineering were found across 49 studies. Games and Project-Based Learning were the most frequently used strategies. Conclusions: Active learning strategies can be combined with others as well as combined with traditional learning activities like lectures in order to foster both lower and higher order thinking skills. Team-Based Learning is one relevant strategy in our Software Engineering Department. Overall, lecturers express Team-Based Learning is a positive addition to courses and would continue deploying it.

Recent studies have highlighted the rise in violence and criminal activities primarilyinvolving firearms. In response to the growing demand for effective firearm detec-tion systems in public safety applications, this paper explores advancements in real-time object detection using Transformer-based models. Building on the RT-DETRarchitecture and its latest version, RT-DETR v2, we introduce improvements such asthe Bidirectional Feature Pyramid Network (BiFPN) for enhanced small object de-tection and dynamic batch processing to maximize computational efficiency and re-source utilization on edge devices like the Nvidia Jetson AGX Xavier for efficient real-time deployment. We also compare the model with state-of-the-art alternatives suchas YOLOv10, demonstrating the superiority of Transformer models in terms of accu-racy and performance. For the comparative study, we used a benchmark proposingthree datasets with challenging conditions. Code and trained models are available athttps://github.com/labt1/GunDetection-RTDETR.

In modern world, continuous data flows that may be analyzed in real-time and are created from multiple sources are referred to as data streams. As data streams become increasingly prevalent across various sectors, preserving the privacy of sensitive information while maintaining data utility is a significant challenge. Due to privacy concerns in most of the organizations, data are shared with third party or the public. In many cases, users are hesitant to provide personal information. This paper proposes a privacy-preservation model for data streams based on condensation-based anonymization. This model uses a condensation mechanism to reduce the granularity of data while ensuring that personally identifiable information and sensitive attributes are effectively anonymized. By leveraging a condensation approach, the model selectively compresses data, allowing for both privacy protection and the retention of essential data patterns for analysis. The condensation method minimizes lost information throughout the anonymization process in an effort to maintain the statistical features of the data. Using generalization and suppression approaches, k-Anonymity obscures identifying information in datasets, making it a useful tool for protecting individual privacy. The proposed model employs hybrid of K- Anonymity algorithm and condensation method to enhance the security and privacy of the data.

This special issue features four invited papers from the 2024 edition of the MSc. and PhD. thesis contests,  organized by CLEI, the Latin American Center for Computing Studies. CLEI's mission includes promoting the  scientific growth of all computing fields across Latin America. To recognize these advancements and foster the region's  graduate programs, CLEI has been hosting the Latin American Master Thesis Contest (CLTM) annually since 1993 and  the Doctoral Thesis Contest (CLTD) since 2016. Both contests are held in conjunction with the Latin American  Computing Conference (CLEI conference).  The 2024 conference took place in Bahía Blanca, Argentina, and received numerous high-quality submissions for both the  MSc. and PhD. contests. Post-conference, the authors of the best works were invited to submit original, unpublished  papers to the CLEIej journal. These submissions went through another evaluation round, adhering to the journal's  standard review process.

This special issue of the CLEI Electronic Journal proudly presents a selected collection of extended and revised papers from the 50th Latin American Computing Conference (CLEI 2024) held at the Universidad Nacional del Sur, SADIO - Bahía Blanca, Argentina, from the 12th to the 16th of August 2024.

For those with movement disabilities, a variety of hands-free mouse replacement systems have been developed, and during the past three decades, numerous advancements have been made. Over the past three decades, numerous authors have proposed alternatives to the mouse for people in the movement with disabilities who have not yet had a fair opportunity to utilise the standard input methods of a personal computer. In camera-based systems, the overhead of using head-mounted devices is reduced by using the web camera as the mouse. Tracking user facial expressions as they are being captured by the camera and accurately translating them into mouse cursor movement and click events are research challenges and opportunities. The current systems can only move the pointer in a slanting manner due to the user's accidental head movements losing the tracked feature. The movement has not yet allowed people with impairments the same chances as others to interact with computers. They have trouble using the input devices on computers due to their mobility problems. Controlling mouse pointer navigation is still difficult, despite the fact that on-screen virtual keyboards may be used to simulate a physical keyboard and speech recognition can be easily utilized to map mouse click events. The development of hands-free mouse replacement technologies has undergone much advancement. There are a number of limitations to mouse replacement systems that use a webcam as a mouse. In order to enable people with movement disabilities to use a standard PC, this research suggests enhancing the ability of camera-based hands-free computing systems to control the mouse cursor by predicting the user's selection of the target item in the GUI-based system using neural network techniques. Using samples where the mouse cursors predicted position values are closer to the user's actual selection region on the computer screen, the system is put to the test, and the anticipated outcome is achieved in every sample.

Current challenges in education lead to the search for strategies to enhance learning and motivate students. Challenge-based learning facilitates meaningful appropriation of knowledge. This article focuses on proposing an innovative educational strategy and a teaching-learning methodology using the I-Tournament tool, aimed at students of the Public Accounting programme, with a focus on its contribution to organizational processes. The research, of an applied and mixed nature, employs a triangular approach with a descriptive scope and a deductive method. The I-Tournament tool was designed as a training and management strategy in academic contexts, initially implemented in SMEs in Bogotá, to address specific organisational needs. The implementation of these continuous training strategies and the exchange of knowledge through technological tools such as I-Tournament promotes continuous training of future professionals, improving their analytical and problem-solving skills, and keeping them up to date in accounting, financial and tax issues.