My research project looks at ways in which we can leverage world knowledge to improve offensive language detection systems both in terms of their downstream performance and explainability.
To sum up, my research interests are: Natural Language Processing, Online Abuse, Knowledge Graphs and, more broadly, Computational Social Sciences.
I chose the CDT because of its cohort-based approach which enriches me every day with different perspectives and new stimuli towards the creation of safer AI models.
Undergraduate Qualification: BSc Economics and Social Sciences, Bocconi University (Italy)
Masters Qualification: MSc in Data Science and Business Analytics, Bocconi University (Italy)
Visiting Student: I spent four months as a visiting student at Dartmouth College (USA)
Work Experience: During my studies, I carried out:
- Research activities (Research assistant at Fondazione Ing. Debenedetti; Visiting student at Bocconi Institute for Data Science and Analytics; Data analyst at Centre for Research on Health and Social Care Management SDA Bocconi)
- Voluntary projects (Omdena and Save the Children)
- Consulting (Junior tech consultant at EY Financial Services).
Conferences/Papers: “Leveraging time-dependent lexical features for offensive language detection” accepted at EvoNLP Workshop at EMNLP 2022.