Data Mining – the process of extracting useful knowledge from large amounts of data – has become incredibly popular both in industry with business intelligence and personalized marketing as well as in government, where new techniques for predicting and responding to terror and crime is being developed and deployed in the USA and England. The commercial successes of Data Mining has many new actors anxious to use these techniques while at the same time having citizens and internet users concerned about the consequences of their wide-spread use.
In the FounData project, our goal is to make the process of Data Mining more transparent and understandable as well as making it easier to critically scrutinize and assess Data Mining analyses and models. The focus of my PhD project is on developing techniques and evaluation methods in response to such questions as:
· How do we address the concerns of people subject to Data Mining models in the development of these models?
· How do we lower the bar of entry into Data Mining without compromising the integrity and quality of the analyses produced?
· How can the results of Data Mining be communicated to stakeholders and decision makers to facilitate decisions on informed grounds?
These questions will be explored in an interdisciplinary context between Data Mining and Human Computer Interaction.
Contact: Simon Enni
Since 1850, Statistics Denmark have been the central authority on Danish statistics. They serve authorities, public institutions, researchers, journalists, students and citizens by curating and making statistical information broadly available in Denmark. In recent years, several other public data platforms have emerged with the aim of supporting transparency, broad access and reuse of data, and ultimately innovation within the digital economy. The parliament publish legislation data on their open data platform an and a long list of municipalities publish datasets on a shared open data platform. In the future, these platforms will play an important role our democracy by ensuring transparency and broad access to the data influencing decision-making
Each of these platforms allow citizens to browse the data through platform-specific data-browsers and access the data with platform-specific application interfaces (APIs). These supports finding a specific dataset and subsequently accessing it through dedicated application or service. However, this is rarely how people interact with information, in particular when they are searching for and seek to compare information across different datasets.
This project explore how data from public data platforms can be made accessible through a combination of interactive data visualisations and data analyses. This allow people to browse the large data repositories through visualisations of the data they contain and compare the data based on shared units, variables, dimensions, statistical properties and common filters, such as time and geography.
Contact: Henrik Korsgaard
