Giorgia Adorni


Postdoctoral researcher
Generative AI models for automatic music creation
Human behavioural sensing and modelling

About Me
I am currently a Postdoctoral Researcher at ISIN, DTI-SUPSI, where I focus on data engineering, human behavioural modelling, and generative AI for music.
My work explores how AI can be used to understand and enhance human experiences, from personalised learning to emotion-aware systems and creative applications like music generation.
My Ph.D. Research
Previously, I completed my PhD at IDSIA USI-SUPSI (Dalle Molle Institute for Artificial Intelligence), where I developed an Intelligent Assessment System to evaluate students’ development of computational thinking and algorithmic skills. My research bridged educational technologies and AI to create adaptive learning environments tailored to individual learners’ needs.
From Art to Technology
My academic journey didn’t start in computer science. I attended an arts high school with a focus on architecture, where creativity and artistic expression were at the forefront. During this time, I also explored other forms of art, such as photography and even tattoo artistry. However, despite my passion for the arts, I found myself increasingly drawn to technology. I was fascinated by how technological advancements can shape the world and improve people's lives, so I took a bold step. I enrolled in Computer Science at the University of Milano-Bicocca (UNIMIB), Italy.
After completing my B.Sc. in Computer Science in 2018, I pursued a Double Master's Degree in Informatics between UNIMIB and Università della Svizzera italiana (USI) in Switzerland, graduating in 2020. This international experience broadened my horizons, giving me a unique perspective on how technology can be applied to different contexts and cultures.
Research Interests
During my PhD, I conducted studies across different Swiss regions and linguistic contexts, using AI to assess and enhance algorithmic thinking in learners of various ages. Today, my research continues to explore:
- Data Engineering and Behavioural Modelling: Developing data-driven solutions for human behaviour sensing, recommendation systems, and personalisation.
- Sentiment and Emotion Recognition: Leveraging advanced ML techniques for affective computing.
- Generative AI for Music: Exploring AI models that create music, merging technology with creativity.
- Educational AI: Investigating how intelligent systems can personalise learning, improve engagement, and foster computational thinking.
Future Goals
I aspire to continue bridging AI with education, creativity, and human behaviour. My goal is to develop AI-driven tools that democratise access to high-quality education, support well-being, and enable creative expression. I'm particularly interested in interdisciplinary collaborations that blend AI with cognitive science, pedagogy, and the arts.
Outside of Academia
Outside of my academic pursuits, I am driven by a desire to make the world a better place. I believe that small acts of kindness can have a big impact, whether it's mentoring a student, lending a hand to a colleague, or simply being there for someone in need.
Music is a big part of my life. Singing and listening to music help me recharge and stay grounded, keeping my creative spirit alive alongside the rigor of scientific work.
Get in touch!

Projects

2021–2022

gansformer-reproducibility-challenge
Replication of the novel Generative Adversarial Transformer.
(Advanced Topics in Machine Learning course @USI)

2019–2020

lstm-for-text-prediction
Implementation of a text generator based on Long Short-Term Memory (LSTM).
(Deep Learning Lab course @USI)

deep-q-network-atari-games
Implementation of a Deep Q-Network agent, based on Mnih et al., 2015.
(Deep Learning Lab course @USI)

ros-turtlesim-controller
A simple ROS node that controls a turtle in turtlesim.
(Robotics course @USI)

mighty-controller
An open loop controller that moves a Thymio in Gazebo.
(Robotics course @USI)

learning-relative-interactions-through-imitation
Learn to perform specific interactions between a marXbot robot and objects in the environment, using the simulator Enki.
(Robotics course @USI)

learning-robot-swarm-controllers
Master’s thesis project that simulates robot swarms for learning communication-aware coordination using imitation learning approaches.
(@USI)

2018–2019

mosquito-disease-spread-rate-prediction
A supervised classification model capable of predicting whether or not WNV virus will be detected in a certain trap every week.
(Data Technology and Machine Learning course @UNIMIB)

viterbi-algorithm-for-hmm
An implementation of the Viterbi Algorithm for training Hidden Markov Models.
(Probabilistic Models for Decisions course @UNIMIB)

likelihood-weighting-sampling
A Python implementation of the likelihood weighting approach for Bayesian Network sampling.
(Probabilistic Models for Decisions course @UNIMIB)

misspelling-corrector
Misspelling corrector based on Hidden Markov Models and Noisy Channel Model.
(Probabilistic Models for Decisions course @UNIMIB)

dct2-compression
DCT2 implementation with Faster-than-light Fourier Transform.
(Methods of Scientific Calculation course @UNIMIB)

comparison-of-cholesky-decomposition-implementations
A study of the implementation of the Choleski method for the resolution of linear systems for sparse, symmetric and positive definite matrices. Comparison based on different open source programming environments and MATLAB implementation.
(Methods of Scientific Calculation course @UNIMIB)

human-disease-network
Human Disease Network: clustering and performance evaluation.
(Data Analytics course @UNIMIB)

2017–2018

neigh
Neighbour Joining algorithm for the creation of Phylogenetic Trees.
(Bioinformatics course @UNIMIB)

chess-recognition
A MATLAB library for recognition of La Settimana Enigmistica chess games.
(Image Processing course @UNIMIB)

learning-personality
Bachelor internship project that uses Neural Networks to predict personality traits from text written in natural language.
(@UNIMIB)

Thesis

PhD thesis:
Towards an intelligent assessment system for evaluating the development of algorithmic thinking skills: An exploratory study in Swiss compulsory schools
GitHub code

Master thesis:
Simulation of robot swarms for learning communication-aware coordination
GitHub code

Bachelor thesis:
Neural networks for learning personality traits from natural language
GitHub code

Research

Recent publications

G. Adorni (2025). Towards an intelligent assessment system for evaluating the development of algorithmic thinking skills: An exploratory study in Swiss compulsory schools. Tech Know Learn.G. Adorni*, A. Piatti, E. Bumbacher, L. Negrini, F. Mondada, D. Assaf, F. Mangili, L. M. Gambardella. (2025). FADE-CTP: A Framework for the Analysis and Design of Educational Computational Thinking Problems. Tech Know Learn.G. Adorni*, I. Artico, A. Piatti, E. Lutz, L. M. Gambardella, L. Negrini, F. Mondada, D. Assaf (2024). Development of algorithmic thinking skills in K-12 education: A comparative study of unplugged and digital assessment instruments. Computers in Human Behavior Reports, vol 15, 100466.G. Adorni*, S. Piatti, V. Karpenko (2024). Virtual CAT: A multi-interface educational platform for algorithmic thinking assessment. SoftwareX, vol 27, 101737.G. Adorni*, A. Piatti. (2024). The virtual CAT: A tool for algorithmic thinking assessment in Swiss compulsory education. arXiv preprint arXiv:2408.01263.S. Corecco, G. Adorni*, L. M. Gambardella. (2023). Proximal Policy Optimization-Based Reinforcement Learning and Hybrid Approaches to Explore the Cross Array Task Optimal Solution. Machine Learning and Knowledge Extraction, vol. 5, no. 4, pp. 1660-1679.G. Adorni, F. Mangili, A. Piatti, C. Bonesana, A. Antonucci. (2023). Rubric-based Learner Modelling via Noisy Gates Bayesian Networks for Computational Thinking Skills Assessment. Journal of Communications Software and Systems, vol. 19, no. 1, pp. 52-64.F. Mangili, G. Adorni, A. Piatti, C. Bonesana, A. Antonucci. (2022). Modelling Assessment Rubrics through Bayesian Networks: a Pragmatic Approach. International Conference on Software, Telecommunications and Computer Networks (SoftCOM), pp. 1-6.A. Antonucci, F. Mangili, C. Bonesana, G. Adorni. (2022). Intelligent Tutoring Systems by Bayesian Nets with Noisy Gates. The International FLAIRS Conference Proceedings, vol. 35.A. Piatti, G. Adorni*, L. El-Hamamsy, L. Negrini, D. Assaf, L. M. Gambardella, F. Mondada. (2022). The CT-cube: A framework for the design and the assessment of computational thinking activities. Computers in Human Behavior Reports, vol 5, 100166.A. Antonucci, F. Mangili, C. Bonesana, G. Adorni. (2021). A New Score for Adaptive Tests in Bayesian and Credal Networks. European Conference on Symbolic and Quantitative Approaches with Uncertainty. Springer, Cham. pp. 399-412.* corresponding author