filippo simini

The last day of #IC2S2 kicks off with an excellent Keynote by @martikagv on urban modelling and leveraging big data to deal with sporadic mobility data. Also introducing scikit-mobility: a library for human mobility analysis in Python scikit-mobility.github.io/scikit-mobilit… @IC2S2

📢 scikit-mobility v1.3.1 is now out! 💡 new version includes a "data module" to download open mobility data and upload new ones to the collection! 🔥 A blog post will show how. Stay tuned! 👉 github.com/scikit-mobilit…

Hello Mobility World! #scikitmobility now on Twitter! github.com/scikit-mobilit…

New paper out! We study the distribution and spatial arrangement of human settlements in the world. We study an (Open!) dataset of satellite imagery at high resolution (10m!) and find two emerging patterns, namely agglomeration and dispersion. nature.com/articles/s4159…

@NatureComms editors' highlights webpage on "AI and machine learning" showcases the 50 best papers recently published in this area. It's also featuring our Deep Gravity paper🏅👉 nature.com/collections/ce… #HumanMobility #AI #DataScience

📢🔥Latest paper on @NatureComms: "A Deep Gravity model for mobility flows generation" w @iPif @luca_msl @GianniBarlacchi - Flows realism improved to up 1000% wrt gravity - Geographically transferable - Flows explanation based on geo variables and POIs 👉nature.com/articles/s4146…

@f_pouchain Hi, this paper discusses the normalization for finite systems arxiv.org/abs/1206.5735 and the python library scikit-mobility has an implementation of the radiation model github.com/scikit-mobilit…

Hey @iPif, I'm trying to implement your radiation mobility model but I'm not sure how to handle the normalization step for finite systems (mentioned in "Human Mobility in a Continuum Approach"): doesn't it generate "impossible" flows (Tij > nj) ?🤔

📢 𝗦𝗶𝗺𝘂𝗹𝗮𝘁𝗲𝗱 𝗣𝗿𝗶𝘃𝗮𝗰𝘆 𝗔𝗻𝗻𝗲𝗮𝗹𝗶𝗻𝗴 (SPA) is an adversarial behavior model for risk assessment in 𝗺𝗼𝗯𝗶𝗹𝗶𝘁𝘆 𝗱𝗮𝘁𝗮. Our new paper out 🔓 on IEEE Transactions on Intelligent Transportation Systems! #datascience #privacy 👉 ieeexplore.ieee.org/document/91998…

Thread) Today I share with you a giant effort to provide an unprecedented estimation of the global cover and geography of human settlements (HSs). We produced a 10 m resolution binary mask outlining HSs and built a model to explain settlement formation arxiv.org/abs/2006.06584

Released our first report on the analysis of anonymized mobile phone data in Italy during lockdown: 1⃣ flow diversity decreases 2⃣ the number of local job markets explodes. preprint 👉 arxiv.org/abs/2004.11278 webpage 👉 sobigdata.eu/covid_report/#… Much more to come. #covid19italy

No need to choose between collective health and individual privacy ncase.me/contact-tracin…

I would like to announce this collection on "Cities as Complex Systems" to appear in PLoS One. Marta @martikagv and myself are guest editors. collections.plos.org/s/cities @PLOSONE #ctcsplos2020 Please feel free to contact us in case you have any question.

COVID-19 Community Mobility Reports buff.ly/2JNdvRb

The proposed methodology is cheaper and more scalable than traditional field surveys, it is privacy-compliant because it is not based on sensitive personal data, and it is significantly easier to implement than methods based on text and sentiment analysis

Analysing the time series of Facebook posts of small businesses before and after an event, we estimate in real time the long-term economic recovery for different types of natural hazards (earthquakes and hurricanes) in different continents

Social media usage reveals recovery of small businesses after natural hazard events: this is what we show in our new paper with @robertey_re and Flavia De Luca in @NatureComms: nature.com/articles/s4146…

Our analysis of the effect of the lockdown on the mobility and contact patterns of Italians is finally public! We are measuring in real-time how mobility and contact patterns changed in Italy during the #COVID19 outbreak through anonymized digital traces: covid19mm.github.io/in-progress/20…

Just advertising three new PhD positions focused on networks dynamics, biodiversity and social data. More soon. recruitingapp-5442.de.umantis.com/Vacancies/575/… recruitingapp-5442.de.umantis.com/Vacancies/576/… recruitingapp-5442.de.umantis.com/Vacancies/577/…

Human mobility workshop @appliedmldays #AMLD2020 by @lucpappalard @GianniBarlacchi @Pellungrobe & @iPif: Scikit-mobility allows you to process mobility data as trajectory & flow dataframes and perform a wide range of analyses as easy as using pandas #skmob bit.ly/skmob_repo