Higher Education Institutions (HEIs) are involved in an evolution to a new model of university called digital university. This model implies not only adopting new technologies but also developing an organizational strategic transformation which includes information, processes, human aspects, and more. Because an organization’s digital maturity correlates with the scope of its digital transformation efforts, this study aims to identify digital transformation initiatives (DTI) taken by HEIs, defining the new processes and technologies used to implement them. The main motivation is to have a real and clear vision of how universities are transforming themselves, discovering the most relevant DTI that they have applied and if they are doing it through an integrated plan aligned with a digital strategy, as recommended by experts. We conducted a Multivocal Literature Review, as methodology research, to include both academic and grey literature in the analysis. Main results show that the DTI implemented are primarily focused on providing a quality and competitive education (24% of 184 DTI from 39 different universities analyzed). Emerging technologies most frequently used are advanced analytics (23%), cloud (20%) and artificial intelligence (16% of total DTI). We conclude that HEIs are in the first steps to digital maturity as only 1 in 4 have a digital strategy and 56% have launched isolated DTI that are not integrated in a plan and do not have a high strategic return value to the organization.
11 January 2020 in European Journal of Electrical Engineering and Computer Science
Regression testing is very important but also a very costly and time-consuming activity that ensures the developers that changes in the application will not bring new errors. Retest all, selection of test cases and prioritization of test cases (TCP) approaches are used to enhance the efficiency and effectiveness in regression testing. While test case selection techniques decrease testing time and cost, it can exclude some critical test cases that can detect the faults. On the other hand, test case prioritization considers all test cases and execute them until resources are exhausted or all test cases are executed, while always focusing on the most important ones. Over the years, machine learning has found wide usage in solving different problems in software engineering. Software development and maintenance problems can be defined as learning problems and machine learning techniques have shown to be very effective in solving these problems. In the range of application of machine learning, machine learning techniques have also found usage in solving the test case prioritization problem. In this paper, we investigate the application of machine learning techniques in test case prioritization. We survey some of the most recent studies made in this field and provide information like techniques of machine learning used in TCP process, metrics used to measure the effectiveness of the proposed methods, data used to define the priority of test cases and some advantages or limitations of application of machine learning in TCP.
31 January 2025 in IEEE Open Journal of Intelligent Transportation Systems
Effective traffic management and control are essential for mitigating congestion and minimizing environmental impacts on road transportation systems. In this paper, we propose a novel approach for traffic modeling that integrates physics-based dynamics with machine learning techniques. Our method leverages Gaussian Processes (GPs) and a multi-class second-order discrete traffic model known as METANET to develop a Physics-Regularized Machine Learning framework. Furthermore, the proposed approach includes for the first time multi-class on/off ramps within the modeling framework, enhancing the realism of the predictive model. We systematically evaluate the performance of the hybrid model across varying dataset sizes to determine optimal data requirements for accurate traffic predictions. Experimental results indicate the improved predictive performance of the proposed approach compared to traditional machine learning and physics-based models. Our findings underscore the potential of Physics-Regularized Machine Learning for enhancing traffic management and control strategies in real-world scenarios.
In this paper, we present a Universal Dependencies (UD) treebank for the Standard Albanian Language (SAL), annotated by expert linguistics supported by information technology professionals. The annotated treebank consists of 24,537 tokens (1,400 sentences) and includes annotation for syntactic dependencies, part-ofspeech tags, morphological features, and lemmas. This treebank represents the largest UD treebank available for SAL. In order to overcome annotation challenges in SAL within the UD framework, we delicately balanced the preservation of the richness of SAL grammar while adapting the UD tagset and addressing unique language-specific features for a unified annotation. We discuss the criteria followed to select the sentences included in the treebank and address the most significant linguistic considerations when adapting the UD framework conform to the grammar of the SAL. Our efforts contribute to the advancement of linguistic analyses and Natural Language Processing (NLP) in the SAL. The treebank will be made available online under an open license so that to provide the possibility for further developments of NLP tools based on the Artificial Intelligence (AI) models for the Albanian language.
ML (Machine Learning) is frequently used in health systems to alert physicians in real time. This helps to take preventive measures, such as predicting a future heart attack. This study presents ML combined with various forms of feature selection to identify heart disease. It includes the analysis of different algorithms such as Decision Tree, Logistic Regression, Support Vector Machine, Random Forest and hybrid models. This results in SVM and RM performing better after applying feature selection for individual ML models. Meanwhile, hybrid cases provide good results if the ensemble is done using a Voting Classifier. Our approach in this paper is based on our study of existing literature and methodologies. We can conclude that, for the used dataset, the Voting Classifier appears to be the most accurate and precise model out of all individual and hybrid classifiers that use feature selection techniques.