[Pycon] [new paper] "Fabio Lamanna" - From the language of the tweet to immigration in cities: Python and its "Language Detectors"

[info a pycon.it]

Title: From the language of the tweet to immigration in cities: Python and its "Language Detectors"
Duration: 45 (includes Q&A)
Q&A Session: 0
Language: en
Type: Talk

Abstract: Oh! Twitter again! In this talk I would like to show you Python under stress working with  350 million tweets written by around 15 million users in 53 world cities in the last five years. We will see how we were be able to extract the most probable mother-tongue of each Twitter user working around their tweeting timeline, thanks to pandas, some language-oriented packages (cld2, LangId) and some statistics.

We then focus on users' spatial distribution in cities, building a special ranking among cities much more friendly in hosting immigrant communities, and others characterized by "ghetto" phenomena.

Thanks to Twitter we developed a methodology that helps in understanding how different communities are spatial integrated/segregated within the major cities in the world.

Prerequisites: nothing

Tags: [u'nlp', u'twitter', u'data-analysis', u'language', u'Social Network Analysis', u'social-media', u'pandas']