Social media predicts urban gentrification
Thu 14 Apr 2016
A team of researchers at the University of Cambridge released a study that shows that social media data can be analyzed to accurately predict urban gentrification. Using economic data from 2010, cross-referenced with social media check-ins on Foursquare and Twitter, and compared to data from 2015, the study found that data analysis predicted which neighborhoods would go through gentrification during that period.
The researchers built an ‘interconnected geo-social network’ of users and venues across London, combining the data of people (using Twitter) and places (using Foursquare). Using data from more than 37,000 users and 42,000 venues, the team created a ‘social diversity’ metric. They then cross-referenced social diversity with the UK Index of Multiple Deprivation (IMD), which measures relative prosperity by neighborhood.
“We’re looking at the social roles and properties of places,” said Desislava Hristova from the University’s Computer Laboratory, and the study’s lead author. “We found that the most socially cohesive and homogenous areas tend to be either very wealthy or very poor, but neighbourhoods with both high social diversity and high deprivation are the ones which are currently undergoing processes of gentrification.”
The team found that the neighborhoods with the highest social diversity and the highest deprivation were primed for gentrification. Of the 32 neighborhoods studied, Hackney had the highest social diversity and second-highest deprivation rating in 2010, and is currently undergoing intense gentrification. House prices are rising more quickly than the London average, crime is falling, and the population is becoming more diverse. Other neighborhoods with similar high social diversity and high economic deprivation are undergoing gentrification as well: Lambeth, Hammersmith, and Greenwich among them.
Social diversity of neighborhoods was broken down into four measures: brokerage, the ability of a place to connect people; serendipity, the prevalence of chance encounters; entropy, the diversity of visits; and homogeneity of visitors. Applying these research methods would be useful for local governments and for urban planning, as the ability to predict which neighborhoods are prone for gentrification provides valuable insight into future needs of that community.