Gentrification predicted by social media data


flickr photo by x6e38 shared under a Creative Commons (BY) license

Gentrification – the process by which a deprived area becomes affluent – can be predicted by data from location-specific social media networks.

The Cambridge University study took data from location-specific social media networks, such as Facebook places and Foursquare, to show that the ‘poor to posh’ process can be identified when high levels of deprivation and social diversity are seen to develop in an area.

Data crunch

Predicting gentrification is more than a spike of social media users tweeting about Radio 4 programmes. It is data analysis of social media check-ins which reveal behaviours derived from the particular places that people go.

37,000 users and 42,000 venues in London were used to build a network of Foursquare places and Twitter visitors, totalling half a million check-ins over a ten-month period.

Researchers quantified the social diversity of various areas by distinguishing between places that:

  • brought together strangers versus those brought together friends
  • attracted diverse individuals as opposed to those attracting ‘regulars’


flickr photo by KylaBorg shared under a Creative Commons (BY) license

This data was then correlated with London wellbeing indicators, such as lower crime rates and rising house prices.

Lead researcher Desislava Hristova:

“We’re looking at the social roles and properties of places. 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.”

In order to measure the social diversity of a given place or neighbourhood, the researchers defined four distinct measures:

  • brokerage (ability of a place to connect people who are otherwise disconnected)
  • serendipity (how a place can induce chance encounters between visitors)
  • entropy (extent to which a place is diverse in respect of its visitors)
  • homogeneity (extent to which visitors to a place are homogenous in characteristics).

In the ‘places’ section, researchers found that particular venues attracted friends where others attracted strangers. Friends were more likely to meet at a fried chicken restaurant, a B&B, a football match or a strip club whereas strangers frequented dumpling restaurants, motels, art museums or gay bars.

Desislava Hristova et al. ‘Measuring Urban Social Diversity Using Interconnected Geo-Social Networks.’