News

New semantic portal gives meaning to parliamentary data

The new tool will help researchers, journalists, and the public to understand and oversee democracy
Illustration of a person speaking on a microphone.
Parliamentary speeches are vital material for research, because they build different political alternatives. Image: Matti Ahlgren/Aalto University.

If you wanted to read all 960,000 speeches held in the Finnish parliament since 1907, it would take seven years and four months as a fulltime job. You’d also have to piece together the digitised, yet unstructured documents from online sources in different formats.

What if all the data were available at your fingertips via a smart web service?

Meet ParliamentSampo, a semantic digital service which has pooled all the speeches and parliamentary networks under a digestible interface. The portal’s name stems from Finnish mythology, where a magical device – Sampo – was bound to bring good fortune to its guardian. In the same way, ParliamentSampo is intended to bear fruit for the wardens of democracy.

‘I think the key contribution of our portal is making these political speeches and networks accessible to the public and providing tools with which researchers and journalists can study the use of parliamentary power in Finland,’ says Eero Hyvönen, professor at Aalto University and director of the Helsinki Centre for Digital Humanities (HELDIG) at the University of Helsinki.

The portal integrates data from several providers, including ministries and parliament, organizing it and enriching it semantically. It provides tools for research, journalism and developers and is an important component in the expanding national semantic web.

‘The traditional web works with links between webpages,’ explains Hyvönen. ‘The Semantic Web works by linking data between webpages. This data has been embedded into a semantic knowledge graph so that a machine can understand the content better and is able to offer more accurate information.’

‘The primary contribution of computational methods is not to generate new information, but to augment our existing research techniques.'

Kimmo Elo, senior researcher at the Centre for Parliamentary Research at University of Turku.

Making the data talk

The portal will be made available to the general public in late 2022, but it’s already helping a handful of researchers make the data talk, diving into discourse, concepts and power.

‘This is unique in the Finnish context because for the first time we have a machine-readable database of all parliamentary speeches,’ says Kimmo Elo, senior researcher at the Centre for Parliamentary Studies at the University of Turku. ‘Parliamentary speeches are especially important because they’re contestations for political alternatives.’

‘If we want to drive progress in computational social sciences and humanities, we need data that is in sync with our methods,’ Elo continues. So far, digital documents have been converted into machine-readable formats on an ad hoc basis by individual researchers.

The new portal broadens the available data, and the semantic approach also offers the potential to challenge and complement existing research. ‘It can inform the way we design our research. Having an organized view of the entirety of parliamentary data allows to look for historical trends, cycles of conceptual and rhetorical evolution, as well as discovering potential blind spots that manual analyses had missed,’ says Elo.

Several research papers to take advantage of ParliamentSampo are already out. For example, Elo and Jenni Karimäki used mixed methods to analyse how environmental policy terminology has changed over the past 60 years in the Finnish Parliament. The findings suggest an increase in the volume and share of environmental policy content, as well as an evolution from a more confined, nature -centred rhetoric towards more holistic and climate change -centred terminology.

The research process demonstrated how computer science can help political scientists. ‘The primary contribution of computational methods is not to generate new information but to augment our existing research techniques,’ explains Elo. ‘In this article, we analysed the data and found clear trends with the chosen terms and concepts, while Jenni – who has a deep understanding of the environmental issues – brought qualitative research techniques to make inferences from the data.’

In addition to magnifying history, the project will ensure a hook-up to current data by incorporating ongoing discussions into its database automatically. Journalists at Helsingin Sanomat were the first ones to put the portal to journalistic use for their live journalism show Black Box.

‘I used the portal’s data in my talk to illuminate the meaning of the plenary session and introduce the most vocal members of parliament,’ explains data journalist Sonia Zaki from Helsingin Sanomat. ‘This list couldn’t have been compiled before due to the sheer amount of manual work it would have entailed. I think’s it’s a valuable tool – especially for political reporters – with which to follow what and how politicians talk in the Parliament.’

Hyvönen’s team also hopes to inspire developers to engage with the open-source tool and find novel ways of using parliamentary data.

Not just semantics

Hyvönen and his colleagues in the Semantic Computing Research Group have been developing the Finnish semantic web for the past two decades under the umbrella of the Sampo model. A series of data services and portals based on the Sampo model link together, and give semantic meaning to, archaeological findings, Finnish war history, biographies, and documents from numerous other fields. Under the hood runs the semantic engine, or the Finnish Ontology and Data Infrastructure, which is a kind of source code for the contents, contexts, and connections of the Finnish Semantic Web.

‘Our work represents a paradigm shift in publishing data in humanities research. From printed texts to online systems, we’re at the third phase where linked data and tools for research are realized – as we have done here,’ says Hyvönen.

He compares the development of a semantic web infrastructure to the impact of physical infrastructure mega-projects. Yet, Hyvönen’s vision for the Sampo model stretches even further.

‘The fourth phase, could be a web where the systems are not only tools, but intelligent agents that help humans in finding and solving research problems.’

Eero Hyvönen

Professor
  • Published:
  • Updated:

Read more news

A man and a woman looking up in a room
Research & Art Published:

How to educate for innovation? New report sheds light on the best practices

The report by Aalto Design Factory dives into the core elements of creating a collaborative learning environment that prioritizes student-centered, real-world problem-solving.
Valkoinen kukka edessä, ihmisiä taustalla
Press releases, Research & Art Published:

Men, Swedish speakers and Master's degree holders most likely to own shares in unlisted companies

The first comprehensive study of Finnish owners of privately held firms found, among other things, that only three per cent of the population own shares in these companies.
Nautakarjaa laitumella.
Press releases Published:

Small reductions to meat production in wealthier countries may help fight climate change, new analysis concludes

Eliminating even a small fraction of current beef production could remove three years’ worth of global fossil fuel emissions
Kolme opiskelijaa ryhmätyön äärellä
Research & Art Published:

The School of Business’ approach to sustainability in business education stands out in Europe

With over 30-year history of research in sustainability topics, we can back up the classroom experience with top level research, Dean Timo Korkeamäki says