Who has suffered the most in the credit crunch? Visualisations that tell the real story for businesses based on the London Gazette official notices

Inspired by the idea of Open Linked Data, the latest demo of the Gazettes’ faceted search released by the TSO Semantic Team fully exploits the Gazettes semantic data and uses semantic technology to answer very real business questions. The proof-of-concept integrates three data sources, published by different organisations and uses five star quality, uniformed format, Linked Data.

TSO has published the London Gazette notice metadata since October 2010. Hosted on TSO’s OpenUp® Platform, the Gazettes Resource Description Framework (RDF) repository provides access to over 120 thousand notices as Linked Data – translating into an impressive 10 million RDF triples, and this continues to grow every day.

The TSO team mashes up the Gazettes notice Linked Data together with Ordnance Survey (OS) geographical information and Companies House data. OS Linked Data such as UK administrative boundary and postcode data is stored in TSO’s OpenUp® Platform and the SPARQL endpoint is made available for public access. Companies House makes all UK company registrations available online via Linked Data Application Programming Interfaces (APIs).This joins the data mix and integrates with company profiles such as ‘nature of business’ to provide an impressive depth of information to trawl.

Powered by Apache Solr, the Gazette faceted search supports both full-text search of the notice content and real-time facet views to cover:
•    Geographic area  
•    Time passing
•    Notice type
•    Nature of company business
 
The search results are visualised in a number of styles as a document list, timeline or map display. The complex levels of information are easily explored by drilling-down into search results. These are also categorised into the hierarchies of notice types, and the nature of business. This enables precise data analysis of areas of interest.

 
“Who has suffered the most in the credit crunch?”
Comparison across geographic boundaries

The Gazettes’ ‘Resolution of winding-up notices’ are created at a significant stage in the Company Insolvency procedure, Figure 1 shows the distribution of these notices all over the country. The density of notices is divided into six different bands, and shows the total number of notices in an area.

The total number of notices is grouped into administrative boundaries at the level of Unitary Authority, District, Metropolitan District and London Borough. The density of notices in a local area is indicated by colour, the geographic boundary is coloured to reflect density of notices in that area – for instance an area in red indicates a large density of companies have wound-up, therefore indicating a greatly affected local area.

Figure 1 ‘Resolution for Winding-up’ notices across the country

By clicking on the red area, it can be seen that the City of Birmingham has the maximum number of this type of notice, at 241.  To see further details about Birmingham City zoom into this area.

Changes in companies winding-up over time

The data visualisation shown in Figure 2 drills down into business sectors and displays how the notices are distributed through time. The hierarchy on the left shows the different types of businesses that wound-up. This is the faceted view of ‘Resolution for winding-up’ notices in Birmingham City split into nature of business (SIC CODE) registered with Companies House.

The top-right timeline widget compares the total number of notices published during a week with the number of selected notice types. In this case it is a ‘Resolution for winding-up’ notice. The bottom-right timeline compares the total number of ‘Resolution for winding-up’ notices with the aggregated number of notices that fall into each business sector, e.g. General Construction, Civil Engineering, and Labour Recruitment.

 

Figure 2  Drilling-down visualisation showing ‘Resolution for winding-up’ notices over time

Geographic map visualisation

There is also a geographic filter, shown by a circle area - centred by postcode within a radius of distance. The faceted search results are displayed on a geographic map widget, as shown in Figure 3. Again, this example focuses on the City of Birmingham and ‘Resolution for Winding-up’ notices.  

The visualisation shows the trading address of companies and the aggregated number of notices for that area on the map widget. The electoral ward boundaries are coloured according to the density of notices.

Figure 3   The geographic search of Gazettes in the local area

Looking ahead

The next stage of this proof-of-concept will focus on integrating the data with other semantic-rich data sets. For example, the Gazettes notices could be correlated with other statistical datasets to provide further meaning to the data, and visualise data so it can be easily processed and understood. The Office of National Statistics data set would be a possible candidate for this.
 
For more information about TSO’s OpenUp® Platform and how it can benefit your organisation contact opendata@tso.co.uk

November 28, 2012, 12:12 pm