A two-year project using "big data" computing techniques to link records of Parliamentary debates with other social data could be "transformative" for political research, the UK academic leading the project has said.
'Digging into Linked Parliamentary Data' is one of 14 international projects to win a share of a £3m Digging into Data Award, an international competition to develop new ways of handling 'big data' in humanities and social science research.
It will develop semantic search tools to analyse databases of political debates in the UK, Canada and the Netherlands from 1800 to the present, enhancing the study of issues such as the history of left-right ideological polarisation in powerful ways, according to its leaders.
The project is headed by the Institute of Historical Research at the School of Advanced Study, University of London with partners including King's College London, the History of Parliament Trust, and the universities of Amsterdam and Toronto.
Dr Jane Winters, reader in digital humanities at the Institute of Historical Research, told UKAuthority.com that UK work will focus on digital versions of the Hansard official record of House of Commons debates and related data such as 'division lists' - how MPs have voted.
Applications could include analysis of key terms to study how left-right ideological polarisation has developed over time in UK politics, and what factors influence this, Winters said. "Is left-right polarisation stronger when a party or politician is in opposition, for example, or in government? And how does it tend to change throughout people's careers?"
Another issue likely to be studied is how discussion of the perennially hot political topic of migration
Has changed over time and through the course of politicians' careers, she said.
"In the case of UK, we can look at how the issue of migration is debated before and after the end of the British Empire - the language used about it and how it evolved; whether there were peaks of interest in Parliament about migration post-war; are they discussed in terms such as a multicultural society, or in economic terms; and how much you can get at that automatically".
As well as using the new tools to mark up data by topic and identify individuials with attributes such as party or gender, the project team wants to link it to wider data on life outside Parliament as held in stores such as Dbpedia, a volunteer-led effort to extract structured information from Wikipedia into databases.
"This opens up fascinating opportunities such as linking statements to politicians' levels of education, or party alliances", Winters said. "For humanities research, this could be transformative."
The new work builds on a previous research project led by King's College London called LIPARM (Linking Parliamentary Records through Metadata). This built a "Parliamentary Metadata Language", or PML, as a standard way of formatting political data.
LIPARM - which ended in January 2013 - built lists of "unique identifiers" for politicians, political job titles, pieces of legislation, Parliamentary sessions and other key data items from Westminster and across the UK's devolved parliaments and assemblies, expressed as codes known as uniform resource identifiers (URIs). Its intention was to help researchers in tracking a politician's career; analysing legislative activity over long periods of time by subject, such as agriculture; and cross-linking parliamentary records with other resources such as newspaper reports; biographical material; multimedia materials from organisations such as the BBC; and court papers.
Support for the Digging into Data Challenge is provided by 10 research funders from four countries including the UK's Arts and Humanities Research Council, the Economic and Social Research Council and academic technology support body Jisc.
Pictured: Bus crossing Westminster Bridge, by Paul Clarke http://paulclarke.com
Digging into Data Challenge: www.diggingintodata.org
Institute of Historical Research: www.history.ac.uk
LIPARM: www.liparm.ac.uk
Dbpedia: dbpedia.org
