Skip to main content

SoBigData Articles

Charting New Frontiers in AI and Data Mining: A SoBigData++ research experience in Pisa

In June 2023, I embarked on an extraordinary journey under the auspices of the SoBigData++ project, a beacon for data-driven innovation in Europe. My destination was the prestigious Institute of Information Technology (ISTI) at the Italian National Research Council in Pisa, Italy. 

This research mission, spanning four weeks, was not just a visit but a collaborative forge where ideas, knowledge, and ambitions were melded to contribute to the future of artificial intelligence (AI) and data mining. In this post, I will summarize my great experience during my research visit. However, it is noteworthy that no blog post will be able to depict wholly what I lived there. For clarity, I will divide my post by weeks and, in the last part of this post, reflect on my experience.


Week 1: The Convergence of Minds


My adventure commenced with enlightening meetings with ISTI's HPC Lab researchers, including Chiara Renso, Raffaele Perego, Franco Nardini, Salvatore Trani, and Emanuele Carlini. These initial interactions were significant and provided me with lots of insights. I shared the essence of my work at UFSC, Brazil, and delved into the groundbreaking research by these HPC Lab members. 

Our dialogues uncovered four fertile grounds for collaboration, ranging from technology-assisted review of legal documents to the avant-garde use of graph-neural networks for unmasking fraud and corruption. 

In parallel with these discussions, I explored the SoBigData++ catalog, identifying datasets that resonated with our collaborative compass. Although this search was very intriguing due to the many interesting available datasets, it indicated no dataset fit the work I planned to develop during my visit.


Week 2: Pivoting with Precision


As the second week unfolded, discussions with Chiara Renso and her Ph.D. Student Chiara Pugliese revealed that the initially identified datasets were, indeed, mismatched with my project's ambitions. However, every challenge is but a doorway to new opportunities. I shifted my focus to a dataset brimming with georeferenced images of
Pisa, unlocking a novel avenue to innovate how we perceive, interact with, and narrate the essence of cityscapes.


Week 3: The Blueprint of Innovation
 

The third week was a deep dive into the architectural phase of our project, where we sketched the blueprint for automatically generating zones of interest within cities. This endeavor was not just about mapping; it was about weaving the fabric of towns into personalized textual stories, using trajectories and georeferenced images as our
threads. This period was marked by vibrant collaborations with Chiara Pugliese, my Brazilian undergrad student Gabriel Rangel, who participated in virtual meetings, and Guido Rochietti, whose expertise in Large Language Models illuminated our path toward automated zone-description generation. 

Spoiler from the future: In the last week of March/2024, we finished the experiments on this research direction, produced an original research article, and submitted it to the European Conference on Machine Learning (ECML). We hope it gets accepted. Fingers crossed!


Week 4: Looking Beyond the Horizon


Other than proceeding with the zone-description research defined in week 3, the final week was marked by strategic planning and establishing an important agreement. In collaboration with Raffaele Perego, we charted the course for advancing Name Entity Resolution through AI, targeting the nuanced recognition of product names in Brazilian fiscal documents. 

The week culminated in a landmark achievement: the formalization of an agreement between ISTI-CNR and UFSC for exchanging graduate students and researchers among these institutions. This pact is a bridge to the future, enabling Brazilian students to partake in research internships in Pisa, thus perpetuating the cycle of innovation and collaboration kindled during my visit.


Reflections on a Journey of Discovery 

Reflecting on my research mission, I am amazed by the fusion of ideas, cultures, and ambitions that characterized my four weeks in Pisa. This expedition was not just about the exchange of knowledge; it was a testament to the unifying power of research and the boundless potential of AI and data mining to tackle some of the most pressing
challenges of our times. 

As I look back on this journey, I am grateful for the SoBigData++ project, ISTI-CNR, and all the brilliant minds with whom I had the privilege to collaborate and who received me so well in Pisa. Together, we have set the stage for a future where AI and data mining enhance our understanding of the world and enrich the human experience in profound and lasting ways. 

Until our paths cross again, I carry with me the memories, insights, and friendships forged in Pisa, a beacon of knowledge that will continue illuminating my research and aspirations in the boundless realm of artificial intelligence and data mining.