Can social media data help us understand our relationship with historic buildings?
Cities are carpeted with pictures, but these pictures are not public art or the work of a subversive street artist. They are the geo-tagged photos continuously being uploaded to sites like Flickr and Twitter. This is creating a new world to explore. A world which could change how we think about the built environment and its heritage.
We know that people engage with heritage (around half the adult English population visit for pleasure a historic town or city in a given year), however, visitor numbers aside, the details of this have always been a bit hazy. Thanks to social media data, how people respond to heritage in the built environment is starting to come into focus. This post looks at some of the emerging findings from research on the connections between London’s photography and its listed buildings (those designated as being of special architectural or historical interest), using data on over 2 million geo-tagged photos collected from the photo sharing website Flickr. The picture below shows part of the data used in the study with the locations of photographs taken in central and inner London (each dot representing a geo-tagged photograph).
Locations of geo-tagged photographs taken in central and inner London
Among the places visible outside the centre are, from the top left going right: London Zoo (the triangular shaped concentration of photos), Camden Market, Kings Cross St Pancras, Upper Street , and the connecting Regents Canal. In the bottom left-hand corner, the Royal Hospital, the Chelsea Embankment, and the Kings Road can also be seen.
In the area where photographic data was obtained (a rectangle bounding the administrative area of London, falling within the M25) there are more than 22,000 listed buildings (654 Grade Is, 1,553 Grade II*s and 20,556 Grade IIs; Grade I being the highest grade of listed building and Grade II the lowest). Higher graded buildings were more likely to have photographs taken near them: 88% of Grade Is had at least one photograph falling within 25m of their centre (as defined by the coordinates given in their list description) as opposed to 76% of Grade II*s and 61% of Grade IIs). The average number of photographs for those that had photographs within this distance was also highest for Grade Is (168 photographs on average), followed by Grade II*s (58) and Grade IIs (42). 
As a measure (admittedly imperfect) of the most photographed Grade I listed building, if we look at the number of photographs taken within 100m of the centre of the building, which seems reasonable given that the highest profile buildings are quite large, and the corresponding number of photographers. Then according to this definition St Paul’s Cathedral (shown below) is the listed building with the highest level of photography surrounding it, both in terms of the number of photographs taken (12,778) and the number of photographers (2,659). 
Geo-tagged photographs around St. Paul’s Cathedral and surroundings
A number of factors will be driving the variations in photography around listed buildings including the building itself, its use, location\surroundings and factors specific to the sample.  Unpacking this is an area of ongoing work. For London overall, looking at larger spatial scales, the city centre is the area with the highest density of listed buildings and it is also the area with the highest concentrations of photographic activity. This does not necessarily imply a direct connection between the two. City centres inevitably have higher levels of pedestrian footfall, and one would therefore expect them to have higher levels of photography.
The picture below shows the density of photographic activity in the sample in central London (the colour scheme, shown right, running from areas of highest concentrations of photographs at the top to areas of lower levels of photography at the bottom). We see that photography is not uniformly distributed. In the east there are a number of concentrations of photography around high profile visitor attractions (many of which are listed). In the west a connected concentration of photography stretches over the river bridges through Westminster and up Whitehall, to Trafalgar Square and Piccadilly Circus, going east to Covent Garden. All of which contain several listed buildings. There are also high concentrations of photography around the British Museum and Buckingham Palace.
Density of photographic activity in the sample in central London (click to enlarge)
At smaller spatial scales there is not necessarily a correlation between areas having a lot of listed buildings and higher levels of photography. A higher density of listed buildings may mean that the buildings are smaller and so are perhaps less likely to be photographed; areas with a high density of listed buildings are often residential and so off the tourist track. Mayfair has a high density of listed buildings, but a low level of photography by comparison. Some of the other places with a high density of listed buildings, graveyards, have residents that are even less likely to take photographs. Which is not to say they are unphotographed. There are, for example, in the sample quite a few photographs of the final home of Thomas Bayes and William Blake: Bunhill fields.
The sample has 50,996 photographers, of whom around half had at least one photograph within 25m of a listed building’s centre. It cannot be guaranteed that these people, or the photographs that they share, are entirely representative of London photography. Indeed they may not be given they are a self-selecting group who have decided to share tagged photos on Flickr. Nevertheless, it would be difficult to obtain this kind of detailed data on photography using more conventional approaches, and this is a substantive proportion of the photographers.
With social media use rising, the data it creates may in the future become increasingly representative of the population and could potentially be considered when making the case for having buildings listed. The recently listed Lloyds building is for example one of the most photographed Grade I listed buildings in the sample. More generally, the data shows that people value the historic built environment even if they do not necessarily visit it as paying visitors. This is illustrated by the concentration of photography in Covent Garden, where, in the 1960s, it was proposed to pull down much of the 18th and 19th century buildings surrounding the market. Picture that.
This article has previously appeared on the English Heritage website. Read the original here.
 The average is skewed by some of the buildings being much more highly photographed than others. If we look at the median number of photographs instead we have for Grade Is (39 photographs), Grade II*s (13) and Grade IIs (7).
 If other grades of building are included then in terms of the number of photographs it becomes the Grade II* footings of the destroyed Cloister and Chapter House (13,554 photographs and 2,739 photographers) which is located outside St Paul’s. Many of these photographs will be counted within the figures for the Cathedral itself. The way buildings are listed may also not necessarily relate exactly to how people perceive them. The Tower of London, for example, consists of multiple listed structures and Tower Bridge is listed in two separate pieces i.e. for the purposes of this analysis it is two listed buildings.
 The uses of listed buildings are often closely bound together with the buildings themselves due to their special character and location. For a discussion of this, based on matching business use data with the location of listed buildings see this report based on a research collaboration between English Heritage and the Heritage Lottery Fund.
 To learn more about this and some of the other issues of analysing social media data see here. There is also, as is common with user-generated content on the web, a skew in the sample in that the 1000 most prolific photographers contribute around half the number of photographs, with a long tail of accounts with a small number of photographs.