Virtual SciDataCon 2021 is FREE! Register now!

Registration link for the session:  https://us02web.zoom.us/meeting/register/tZUsc-igrTwjGdSz0zQj_-1eebiNsCs8TCta

26 October 2021, 11:00 - 12:30 UTC

Data science is an opportunity for boosting social progress, and data analysis tools are triggering new services with a clear impact on our daily life.  SoBigData RI (http://www.sobigdata.eu) is a multi-disciplinary research infrastructure aimed at using social mining and big data to understand the complexity of our contemporary, globally-interconnected society.

The SoBigdata RI’s service platform empowers researchers for the design and execution of large-scale social mining experiments. Pushing the FAIR (findable, accessible, Interoperable, responsible) and FACT (Fair, Accountable, Confidential and Transparent) principles, the RI renders social mining experiments more efficiently designed, and repeatable by leveraging concrete tools that operationalize ethics, incorporating values and norms for privacy, fairness, transparency and pluralism; also touching upon how data science helps us to make more informed choices, underlining the need to achieve collective intelligence without compromising the rights of individuals.

The format of the session is a mix of research and practice presentations where the SoBigData++ project is presented in all its parts:

• SoBigData++: sustainable Platform and ecosystem for Ethical Social Mining (10 min.)

This talk introduces SoBigData project with the aim of putting in context the participants presenting the main objectives of the project and the consortium of experts involved working on the vertical contextes:  Societal Debates and Online Misinformation, Sustainable Cities for Citizens, Demography, Economics & Finance 2.0, Migration Studies, Sports Data Science, Social Impact of Artificial Intelligence and Explainable Machine Learning. Part of this presentation will be the description of an ethical approach to data science which is a pillar of the SoBigData++ project. (Speaker: Roberto Trasarti)​

• SoBigData RI - Facilities, Services & Events (10 min. + 5 min. discussion)

With the second presentation all the facilities and services of SoBigData++ are presented, such as the e-infrastructure, the catalogue, the computational resources, the access to transnational access programme (TNA) and future open events. (Speaker: Valerio Grossi)

• Research highlights (20 min. + 10 discussion)

In the research highlights some of the newest research results supported by SoBigData++ community will be presented to the audience:

Paper 1 - A deep gravity model for mobility flows generation. (to appear with open access in Nature Communications)

Paper 2 - GLocalX -- From Local to Global Explanations of Black Box AI Models. (https://arxiv.org/abs/2101.07685) 

(Speaker: Luca Pappalardo)

• Hands-on computational engine and SoBigData libraries (30 min. + 5 min. discussion)

​Finally the practice session on how to access and use resources provided by SoBigData++. It will be a small tutorial showing how to use the e-infrastructure cloud services to:

• Interactively code a new algorithm or use an existing one from the catalogue 
• Upload a new dataset or use an existing one from the catalogue
• Take advantage of the cutting edge libraries created by our researchers (Tool for the estimation of retweet cascade graphs, DCI-based cross-language text classification, NDlib - Network diffusion library, Scikit-mobility, Anonymization and risk assessment library, etc.).
• ALIR: Distributed training of word embeddings - method for scalable training of word embeddings
• Exploit the full power of the SoBigData++ computational engine for new research

(Speaker: Giulio Rossetti)

Each presentation has time at the end for discussing with the audience.