ISI Foundation presents two papers at KDD2020

The annual KDD conference is the premier interdisciplinary conference bringing together researchers and practitioners from data science, data mining, knowledge discovery, large-scale data analytics, and big data.  The 26th edition of KDD is being held as a virtual conference, with online talks and panels, from 23rd to 27th August 2020. ISI Foundation is joining the conference with two new papers: “Explainable Classification of Brain Networks via Contrast Subgraphs” and “Learning Opinion Dynamics From Social Traces”.

In the first one, Francesco Bonchi (ISI Foundation), Tommaso Lanciano (Sapienza University, Rome, Italy) and Aristides Gionis (KTH, Sweden) target an important problem in neuroscience: the identification of connection patterns that might be associated to specific cognitive phenotypes or mental dysfunctions. They propose a simple, elegant and effective model, based on extracting contrast subgraph, that has great accuracy and excellent interpretability. The paper will be presented  Tuesday at 7 PM and Wednesday at 7 PM (Pacific Standard Time).

“Learning  Opinion Dynamics From Social Traces” is a full ISI Foundation teamwork,  in which Corrado Monti, Gianmarco De Francisci Morales and Francesco  Bonchi propose an inference mechanism for fitting a generative, agent-like model of opinion dynamics to real-world social traces: a  model that retains the benefits of agent-based ones, while adding the ability to perform model selection and hypothesis testing on real data.  It will be presented in the Poster Q&A Sessions (Tuesday and Wednesday at 6 PM) and with an Oral Presentation (Wednesday at 10 AM).

Explainable  Classification of Brain Networks via Contrast Subgraphs”, KDD2020  Poster Sessions (Tuesday at 7 PM, Wednesday at 7 PM)
Full article
Video

Learning  Opinion Dynamics From Social Traces”, KDD2020 Oral Presentation  (Wednesday at 10 AM), Poster Sessions (Tuesday at 6 PM, Wednesday at 6  PM)
Full article
Video