Data on the volume of local internet content (LIC). They are based on the UK Web Archive (https://www.webarchive.org.uk/), which draws data from the Internet Archive, and measure the volume of archived online content of local interest at the MSOA/IZ level during the 2002-2012 period.

Please see the PDF data item listed here, for full details of the dataset, and references. The following is a summary of the PDF:

Each entry of the JISC UK Web Domain Dataset (https://data.webarchive.org.uk/opendata/ukwa.ds.2/) (Jackson, 2017), which is a subset of the Internet Archive and curated by the British Library and includes all the archived webpages under the .uk top level domain, contains a timestamp, the URL of the archived website as well as the British postcode found on each site.

We can use this dataset to derive a measure for how much local internet content (hereafter LIC, Tranos & Stich 2019) there exists across the UK. However, the dataset includes websites with differing geographic reach; some websites may refer to single postcode, while others may refer to several postcodes all over the UK. As described in Tranos and Stich, 2019, this poses a problem when trying to ascertain whether localised internet content is a driver of online behaviour.

We thus need a way to discount websites that have less of a local focus. The underlying idea is that websites that have a high geographic dispersion are less “local”. To compute the geographic dispersion of a websites’ set of postcodes p we calculate the Radius of Gyration (RG) of p in kilometres. A website with a high RG will be of national interest, while a website with a low RG will have a very local geographic presence.

As local geographical units we utilise the Middle Layer Super Output Areas (MSOA) for England and the Intermediate Zones (IZ) for Scotland. For each MSOA/IZ with a set W of archived websites we calculate yearly measures of the volume of LIC.