This work aims to investigate the available platforms for digital archiving, including open-source solutions, and the implementation of such a solution at Transilvania University of Brasov (UniTBv). The initiative is aligned both with national plans for the development of the digital infrastructure for open science - in particular through the university's participation in the National Open Science Cloud Initiative (RO-NOSCI) - and with European guidelines for promoting transparency in research in South East Europe, reflected in the NI4OS-Europe project. At the national level, there is a lack of a coherent policy on the promotion and adoption of open access practices to research results, which contributes to their marginal presence in the public space. Facilitating open access to scientific information could have a significant impact on the academic community and civil society, stimulating inter-institutional cooperation and enhancing the dissemination of knowledge to citizens.

The metal-oxide-semiconductor field-effect-transistor (MOSFET) is the most used transistor nowadays in both digital and analog applications. With the need for low-power circuits and some analog computation functions, the subthreshold (low current) regime of the MOSFET is also used. Therefore, in modelling one must account for two current domains, over threshold voltage and subthreshold, with two different drain current versus gate voltage formulas. Also, the point of transition from one formula to the other must be determined, therefore a longer calculation time is needed. It is useful to determine a unified model and corresponding formula for MOSFET drain current, so the calculation does not depend on the above or subthreshold value of the gate voltage. This is determined in this work together with the extraction of the MOSFET parameters in a practical case.

Digital systems are becoming increasingly prevalent making a qualified cybersecurity workforce and a foundational "cybersecurity culture" essential in all professional fields. Traditional higher education finds it challenging to adapt curricula rapidly to the practical demands of the cybersecurity field, especially for non-specialist students and not only. This program addresses the gap by providing foundational, hands-on cybersecurity education to students from diverse fields of study, building upon existing knowledge about the necessity for cross-disciplinary cybersecurity competence.

The main goal of the current work is to build the “bridge” between theoretical knowledge and practical skills required by the current cybersecurity practices based on a hybrid approach. The initiative aims to train mixed-discipline students in vital cybersecurity activities, developing a broad cybersecurity culture and preparing them for roles focused on delivering security solutions. The project also seeks to prove the effectiveness of an interactive, practical training platform for this mixed student body.

The anticipated outcome is the successful development of content and training of students in practical cybersecurity solution delivery, in alignment with Equity, Diversity and Inclusion (EDI) principles. It is expected for participants to achieve a measurable increase understanding, knowledge and application of concepts like SIEM and Ethical Hacking, stepping from theoretical background to practical deployment and auditing of security solutions. Additionally, the project anticipates the design of a scalable, validated model for cross-disciplinary cybersecurity education that can be replicated in other institutions and fields of study, fostering a campus-wide cybersecurity culture. By the end of the training program, students are invited to offer cybersecurity services to local community organizations (LCOs).

In the last decades Machine Learning models have become increasingly used in different areas of interest, leading to the increase of computational and energy resources. In this context, sustainability has become a global imperative, and Green AI is currently used to encapsulate the concept of development of high-performing and eco-friendly models.

We propose in this paper a function for the selection of the best prediction model from a model library, based on a criterion which combines accuracy, latency, and carbon footprint.

Recommendation systems and sentiment analysis are two of the most common tasks that require high resource consumption. The purpose of this study is to create a system that is based on these two applications with a user-friendly desktop interface. Unlike a traditional system that involves running a single model for each functionality, this one will have to choose the optimal model based on the real-time values of carbon emissions intensities from Romania region obtained from an Electricity Maps API and other criteria based on user input, also taking into consideration the type of device on which the model is running. This approach represents a step in the emerging direction of Green AI paradigm, which also represents the motivation of this study.

The proposed system is composed of two main components: (1) image-based recommendation system and sentiment analysis. The recommendation system uses the Fashion Product Images (Small) dataset from Kaggle platform. The candidate pre-trained CNN (Convolutional Neural Networks) models are: ResNet50, InceptionV3 and MobileNetV2. The score function for the model selection takes into consideration the following aspects: the accuracy and latency of K-Nearest Neighbors algorithm for image recommendation of the entire dataset, the documented latency of each CNN model, the resolution of uploaded image by user, the emission factor score and the device where the model runs. The image resolution criterion offers a “bonus” if the image is close to the model target or a “penalty” otherwise. The emission factor score value adjusts the model score based on the emission factor value from the API. Higher values disadvantage complex models, while simpler models gain preference. This logic applies to the sentiment analysis component, too, where the utilized models are BERT (Bidirectional Encoder Representations from Transformers) and VADER (Valence Aware Dictionary and sEntiment Reasoner), here a particular criterion being the text length.

Few results of this study are as following: In the recommendation system, it was observed that for large dimension images, if the emission factor from the API exceeds the value of 360 gCO2/kWh, the selected model is MobileNetV2 in support of sustainability and for the sentiment analysis component, if the length of text exceeds 50 characters and the emissions value are over 300 gCO2/kWh, VADER is the selected model.

According to the official documentation of the utilized models, the results align with expectations, reflecting each model’s complexity and dimension. These key characteristics along with CO2 emission data, guide the selection of the appropriate model based on the platform used to run the app and user input.