Description: Learning Pulse explored whether using a machine learning approach on multimodal data such as heart rate, step count, weather condition and learning activity can be used to predict learning performance in self-regulated learning settings. The experiment lasted eight weeks involving nine PhD students as participants wearing a Fitbit HR wristband and having the applications on their computer recorded during their learning activities. A software infrastructure for collecting multimodal learning experiences from different sources was implemented based on the xAPI standard and stored in a cloud-based Learning Record Store. As part of this infrastructure a Data Processing Application was developed to pre-process, analyse and generate predictions to provide feedback to the users about their learning performance. The participants of the experiment were asked to rate their learning experience through an Activity Rating Tool indicating their perceived level of productivity, stress, challenge and abilities. These self-reported performance indicators were used as markers to train a Linear Mixed Effect Model to generate learner-specific predictions of the learning performance.
My role : self-directed master thesis project. I was responsible for modelling the problem, designing a representation of the data collected, define the hypotheses, develop the Activity Rating Tool for self reports, exstensive analysis of the data collected.
Outputs : master thesis, workshop paper at CROSS-LAK'16, full paper at LAK'17
Description: IBM Extreme Blue was an outstanding experience which took place during the 3 summer months at IBM Amsterdam. The classic seetting of the internship is the following: a group of 4 high level students with mixed technical and business backgrounds are given a technological challenge from one real IBM's client. The aims first to get a large quantity of ideas, which are selected and combined iteratively with the client to get a final idea which addresses the client's challenge.
In our case, the client assigned to the team was the telecomunication company Liberty Global; the selected idea was the development of an Autonomous Emergency Hotspot Drone, eventually named “the Ziggo Drone”, an out of the box technological solution, whose mission was to differentiate Liberty Global customer experience and bring about innovation in the telecomunication industry.
The project consisted in writing a Business Case, inclusive on an extesive description of the technological solution, and developing a prototype of the drone, as a proof-of-concept. The final product was fully working hotspot drone which was able to provide WiFi and make live-streaming from the on-board camera. The project concluded with the 2015 European Extreme Blue Expo, held at B.Amsterdam, where the Ziggo Drone was awarded one of the best 8 projects among 20 others.
My role: technology expert, envisioned and described the full technological solution of the Ziggo Drone; installed the on-board computer (Raspberry PI with camera), coordinated the work of the team.
Description: Semantic Video Tool (SVT), project assignment presentation for the course on Information Retrieval and Text Mining of the Master course in Artificial Intelligence o Maastricht University. The assignment consists in the realisation of webbased application entitled Semantic Video Tool (SVT for short). Given the transcripts of more than 1300 videos web-crawled from TEDTalk, SVT is able to analyse the content of the videos dynamically through a number of natural language processing and text-mining operations. For each video it extracts keywords and it generates automatic summaries and valuable insights; it also works as a search and a recommendation engine.
My role: project carried out individually.
Description: This project took place in the second semester of the first year of the master Artificial Intelligence at Maastricht University. It is related to a R&D project carried out by Maastricht University and funded by the European Commission. The goal of the project was to help the European Commission determining the effectiveness of the investments in Nanotechnology projects through the analysis of publication data in the field of nanotechnology. Using a variety of data mining techniques such as information visualisation and semantic technologies, a data set of 190,692 publications has been analysed. Several models were trained to predict, the number of citation for a publication, the success of certain keyworkds or to visualise the use of certain keyworkds in specific geographic zones. The project, divided in 10 tasks to be performed during six months, leaded to a number of results fully reported in the final report. The results of the analyisis can be used by decision makers to better understand on what to invest.
My role: designing and implementing the static and dynamic visualisations of the dataset; designing the Ontology of the publications and finding a way to filter the publication talking about the Rare Heart materials.
Description: During the master studies in Artificial Intelligence of Maastricht University, in the first semester of the first year took place a Research Project. The project team, composed by four people, was asked to create an AI for the 2014 MIT Battlecode competition.
The team worked for 6 months in developing the different aspects of the multi-agent AI, taking care of the Micro and Macro management, the strategy, the communication among the agents, the task dispatching and so on.
My role: developing the Opponent Modelling module to make the AI adaptive, contributing in the harvesting strategy.
Description: The Bachelor in Informatics and Digital Communications was written at the Technology Enhanced Learning Lab of the Computer Science Depertment the University "Aldo Moro" of Bari (Italy). There I investigated the role of Learning Analytics in Distance Learning Environments. Starting from an analysis of distance education and technology enhanced learning and of formative assessment, in the bachelor dissertation I theorize a Model on how to leverage Learning Analyics to improve the learning and the teaching processes. Hence, I envision two Dashboards both for the teacher and the learner.
My role: project carried out individually.