City attractiveness seen through Twitter
summary
Method measures city attractiveness by tracking geolocated tweets as mobility proxies and produces three rankings (city-to-world, world-to-city, city-to-city) plus a city interaction network; implementation uses MongoDB + Twitter streaming API and a 100 km grid to compute coverage and normalized radius; dataset is ~20 million tweets and 500k users over ~1,000 days across 58 cities; validation shows >0.9 correlation with phone data and stable results from 200–550 sampled users, data sharing blocked by Twitter TOS.
AI-written summary · about
chapters (19)
- 0:00Twitter-based city attractiveness study
- 1:08Twitter data for human mobility
- 5:01Geolocated tweets from Twitter API
- 6:3258 global cities analyzed
- 9:32Visitor travel after arrival
- 13:37Travel patterns across cities
- 16:18Residents vs tourists travel
- 18:33Visited locations coverage
- 20:29Communities in travel patterns
- 23:08Travel to global landmarks
- 25:15Tweet data to mobility insights
- 28:40Global travel patterns surprises
- 31:18Data reliability validation
- 32:47Landmarks by visitor origins
- 35:06Q&A: Data sharing and API limits
- 36:18Q&A: API restrictions handling
- 37:54Q&A: Retweets and mentions
- 38:17Q&A: Geolocated tweets percentage
- 40:17Q&A: City radius criteria
related talks
16mGeostreams
15mData-Driven Cities
30mAnalyzing COVID-19 Protest Movements: A Multidimensional Approach Using Geo-Social Media Data
21mVisualizing data with JavaScript at Twitter
40mHow to measure a city
26mSocial Networks and Protest Participation: Evidence from 130 Million Twitter Users
41mLarge Road-Network analysis with GraphX
11mThe Value of Social (for) TV