GeoFirm All-island

Cluster mapping for the all-island economy

GeoFirm All-island uses geo-coded, firm-level data for Ireland and Northern Ireland to identify clusters of business activities across the island. The website extends existing capabilities for mapping the spatial distribution of industries, providing an all-island and sectoral perspective on potential development and collaboration possibilities.

GeoFirm All-island has been developed by Queen's University Belfast and the Enterprise Research Centre in partnership with Data City Innovations Ltd. The project has been supported by the Department for the Economy (NI).

In GeoFirm All-island sectors are defined not by traditional standard industrial classifications, often unrelated to emerging sectors, but by proprietary Real-Time Industrial Classifications or RTICs. RTICs are based on algorithmic allocation of firms to industries such as immersive technologies, clean tech, or MedTech, using web-scraped data and machine learning.

Using GeoFirm All-island is designed to be straightforward. You can:

Select specific RTICs or sectors and profile the spatial distribution of firms within the RTIC across Ireland and Northern Ireland.
Obtain overall statistics on the size of each cluster, number of firms etc.
Use download options to focus on different locations, capture and download cluster maps and download aggregate statistics.

Configuration

RTICs

Clustering settings

Clustering algorithm GMM HDBSCAN

GMM clustering parameters

Number of clusters

Exclude outlying %

Exclude clusters below % of total

HDBSCAN clustering parameters

Min points

Min cluster size

Clusters map

Map settings

Toggle heatmap Toggle clusters Show companies not in clusters Force cluster outline display

Heatmap parameters

Radius

Point intensity

A Technical Report providing the background and more detail on the clustering approach adopted in GeoFirm All-Island is available at: Link to report here

We welcome feedback on the current version of the GeoFirm All-island toolkit. If you have any comments or want to know more about the tool or its underlying data, please get in touch with Stephen Roper (Stephen.roper@wbs.ac.uk) or Nola Hewitt (nm.hewitt@qub.ac.uk).