I am currently R&D software engineer at Expert.ai. Since November 2017, I am a PhD student in Smart Computing. I received my Bachelor’s Degree in Computer Science (Ingegneria Informatica) at University of Florence. In April 2017, I obtained the Master’s Degree with honors in Computer and Automation Engineering at University of Siena defending the thesis called “Information Extraction by Learning Deep Architectures from Constraints”.
Research Interests
I have always been interested in Machine Learning and Optimization problems. In the last few years I mostly worked on Natural Language Processing. The goal of my research project is in Conversational Agents applications, which involve Information Extraction and Language Generation. I am currently working in Text Generation, with particular focus on poem generation. My interests include also, word representations, recurrent neural networks, transfer and few-shot Learning.
Projects
:: VULGARIS
Analysis of Italian Diachronic Language Varieties.:: POEMGEN
Language models to generate tercets or verses resembling the style of Dante and Petrarca poets.:: MACHINE LEARNING MANUAL
An introduction to some machine learning tools within the most common development environments.
Awards
:: 1st place Hackathon Mediaset Infinity.
Predicting user churn on the online streaming platform.
:: 1st place Hackathon Soccer Data Challenge.
Predicting football players modern roles and analysis of teams’s best schemes.
Links
Personal Web Page
Github
Linkedin Profile
Twitter
Publications
List of publications on Google Scholar and Researchgate.
Language Processing Technologies Class – Seminars
Deep Learning in NLP – lecture 1
Language Generation – lecture 2
Entity and Relation Extraction – lecture 2
Andrea on this site
- Villanova2: open European language models from Villanova.AI
- Multi-SFT: Permissively-Licensed Multilingual Instruction-Following
- Kamyar Zeinalipour wins Best Fooling AI Agent at Turing Hotel Italy, CLiC-it 2026
- Challenging the Abilities of Large Language Models in Italian: a Community Initiative
- PharmaQA.IT: an Italian dataset for Q&A in the pharmaceutical domain
- ECWCA, Educational Crossword Clues Answering: A CALAMITA Challenge
Show 15 more
- Clue-Instruct: Text-Based Clue Generation for Educational Crossword Puzzles
- The WebCrow French Crossword Solver
- TAILOR WebCrow Workshop
- Language Models for Text Understanding and Generation
- [Nov 4th 2020] LabMeeting: Introduction to Transformers
- [Jan 29th 2020] LabMeeting: Learning in Text Streams, Discovery and Disambiguation of Entity and Relation Instances
- [Nov 13th 2019] LabMeeting: Neural Text Degeneration
- La Nazione, Jul 25th 2019: Dante vs Petrarca
- Neural Poetry: Learning to Generate Poems using Syllables
- First Place at SoBigData Soccer Data Challenge
- [Oct 18th 2018] LabMeeting: Machine and Poetry: can computers learn to compose verses and tercets?
- First Place at Hackathon Infinity
- [ Jun 14th 2018] Lab Meeting: A Character-Aware Neural Model to Learn Word and Context Representations
- [March 22nd 2018] LabMeeting: Dialogue Generation from Structured Knowledge
- Neural Networks for Beginners. A fast implementation in Matlab, Torch, TensorFlow
Lisa Graziani
Giuseppe Marra
Vincenzo Laveglia
Dario Zanca
Alessandro Rossi
Alberto Rossi