CROS

 

 

Eurostat is organising the fourth round of the European Big Data Hackathon from 2 to 7 March 2023 (including the presentation by the winners at the NTTS) as an in-person meeting in Brussels.

The European Big Data Hackathon is organised by the European Commission (Eurostat) and will gather teams from all over Europe to compete for the best solution to a statistical challenge. The teams develop innovative approaches, applications and data products combining official statistics and big data that help to answer pressing EU policy and/or statistical questions. In 2023, the topic of the hackathon will focus on the use of financial transactions' data. 

For any points relating to logistics and questions on the hackathon procedures, please contact our functional mailbox: ESTAT-EU-BD-HACKATHON@ec.europa.eu

European Big Data Hackathon 2023

 The hackathon pitching session recordings are available here: https://webcast.ec.europa.eu/hackathon-pitching-20230306  

 For presentations of the hackathon teams, see here  

 

Eurostat organised the fourth round of the European Big Data Hackathon from  2 to 7 March 2023 (including the presentation by the winners at the NTTS), as an in-person meeting in Brussels. The purpose of  the 2023 hackathon was to foster expertise in using Big Data sources relating to financial transactions for producing innovative ideas for products and tools relevant for the EU policies. 

The European Big Data Hackathon is organised by the European Commission (Eurostat) every two years and gathers teams from all over Europe to compete for the best solution to a statistical challenge. The teams develop innovative approaches, applications and data products combining official statistics and big data that can help to answer pressing EU policy and/or statistical questions.

For points relating to logistics and/or questions on hackathon procedures, please contact our functional mailbox: ESTAT-EU-BD-HACKATHON@ec.europa.eu

Dates

The Hackathon took off on 2 March 2023 with the announcement of the challenge, and closed on 6 March when the teams presented their solutions. The winners were awarded on 7 March 2023, at the first day of the NTTS 2023, where they had the opportunity to present their solutions more in detail. 

Participation in the Hackathon

Twenty-four hackathon teams of National Statistical Institutes’ experts and four teams of students from EMOS network universities competed in the 2023 European Big Data Hackathon. 



 

 

 

 

 

 

 

 

 

 

 

 

  

Data sources

For the purposes of the 2023 European Big Data Hackathon, Eurostat provided to participants a dataset containing fully anonymised daily financial transactions data made available to it by the data's owner, Fable Data Limited.

Data challenge

In the 2023 European Big Data Hackathon, the teams were challenged to design an early warning system relevant for EU policy makers. It was up to the team to choose their own policy question that they wished to answer with their application.

 

2023 European Big Data Hackathon

DATA CHALLENGE

Devise a data analytics tool (with data visualisation features) that: 

  • thoroughly examines   the structure and patterns of the provided granular data on card payments, as collected by Fable Data (and/or available in other datasets used by your team during the hackathon) and 
  • signals deviations from those. 

Based on the data analytics tool, build an early warning system for EU policy makers. 

You are free to: 

  • select a set of indicators, which will – in your view – use the potential of the dataset to the greatest possible extent and are relevant from the perspective of EU policies, and 
  • define (a range of) ‘alert triggers’. 

Please note that, as a first step, you should analyse the data quality, in particular its consistency over time, and might need to choose a subset of data on which to base your data analytics tool. 

Please document and justify your choices. 

Hackathon winners

All solutions proposed by hackathon teams were evaluated based on five criteria: relevance for specific policy use(s), methodological soundness, communication, innovative approach and replicability.  Following the evaluation process, seven solutions proposed by hackathon teams were awarded with trophies. In the 2023 European Big Data Hackathon, for the first time, a special EMOS Prize for an application designed by an EMOS team was awarded.

For the seven teams that won the 2023 European Big Data Hackathon, and information on the winning applications, see below:

Place/Award Hackathon team Winning applications
1st Prize Destatis Team 2 'Subwatch’ on tracking the ‘subscription economy’’; based on credit card data, the team’s application tracks monthly spending on subscription services (not (yet) captured by statistical surveys), for: 1) indications of a looming economic crisis (subscriptions cancellations by households as sign of lost optimism in the economy and its outlook) and 2) the state of the digital economy in a detailed, timely manner
2nd Prize Destatis Team 3 ‘Twinning Europe’, to support regional policy makers by identifying statistical twin regions’; by comparing a region's performance with the performance of its ‘statistical twins’, the application identifies meaningful early warning indicators  and alert triggers for different policy areas, such as business performance, health, mobility, and other.
3rd Prize Statistics Iceland ‘CONSUME’ (Consumer Spending and Economic Monitoring Engine); a dashboard helping policymakers, economists and other users of official statistics to monitor changes in consumer behaviour and economic trends by tracking spending behaviour by various characteristic over time and providing interactive time-series and geostatistical plots. 
EMOS Prize EMOS University of  Porto ‘Using card transactions data for early detection of disease outbreaks’; an early warning system for detecting pandemic outbreaks, which leverages credit card transaction data from drug stores and pharmacies. It uses time series techniques to detect significant increases from the expected spending and transaction patterns. Combined with other data sources, it should help public officials to take preventive measures to reduce the impact of infectious diseases on public health.
4th Prize Statistics Denmark ‘EUquALL: An economy that works for all’; dashboard to be used for monitoring consumption equality trends in Spain (and Germany) between several groups (age, region, gender) over time
5th Prize INSEE ‘4C: Control credit card consumption’; interactive data visualisation of card payments that helps detect changes in consumer spending for a wide range of sectors
6th Prize Statistics Poland ‘Transactional Data Monitoring Dashboard with Triggered Alerts for EU Policy Makers’; the statistical tool monitoring in real-time the use of financial transactions by citizens and foreigners in EU member states and presenting a large number of defined KPI indicators in multidimensional views. 

 

Four winning teams - the three best NSI teams  and the EMOS team - were invited to present their hackathon solutions at the 2023 NTTS. Below you can look up the NTTS slides of the winning teams. To watch the NTTS 2023 Hackathon Awards Ceremony, with the presentations by the four Hackathon winners, see the NTTS 2023 recording [from 10:32 to 11:55]:  https://webcast.ec.europa.eu/ntts2023-day-1-gasp-20230307

 

Team Flag Link to the NTTS slides of the winning teams
Destatis Team 2 SubWatch
Destatis Team 3

TWINNING EUROPE
Statistics Iceland CONSUME Consumer Spending and Economic Monitoring Engine
EMOS University of Porto Using bank card transactions at pharmacies for early detection of disease outbreaks

Pitching Slides

The teams creatively approached the data challenge and put forward a number of novel ideas and solutions for the early warning system based on financial transactions data. To explore the solutions proposed by the hackathon teams, see the teams' slides from the hackathon pitching of 6 March 2023.

 
Team Flag Title of the presentation Link to slides
Statistics Sweden Change indicators of consumer mobility in single markets 01_NSI_SE_EUBD23_pitching slides
Statistics Iceland Consumer Spending and Economic Monitoring Engine 02_NSI_IS_EUBD23_pitching slides
Destatis Team 1 TRIP: Travel Rapid Information Platform 03_NSI_DE1_EUBD23_pitching slides
EMOS University of Pisa Fair Competition? A Prediction Based Alert Mechanism 04_EMOS_Pisa_EUBD23_pitching slides
EMOS University of Athens DELTA  Digital paymEnts officiaL sTAtistic 05_EMOS_Athens_EUBD23_pitching slides
EMOS University of Porto Using bank card transactions at pharmacies for early detection of disease outbreaks 06_EMOS_Porto_EUBD23_pitching slides
EMOS University of Madrid ALCOHOL EXPENDITURE AMONG YOUTH 07_EMOS_Madrid_EUBD23_pitching slides
Statistics Estonia Can German credit card transaction data be used for estimating poverty? 08_NSI_EE_EUBD23_pitching slides
Statistics Norway A Tool for Measuring Seasonal Patterns in Tourism Expenditure 09_NSI_NO_EUBD23_pitching slides
Statistics Poland Transactional Data Monitoring Dashboard with Triggered Alerts for EU Policy Makers 10_NSI_PL_EUBD23_pitching slides
Statistics Belgium Early detection of major changes in market communities 11_NSI_BE_EUBD23_pitching slides
Destatis Team 2 SubWatch 12_NSI_DE2_EUBD23_pitching slides
Statistics Lithuania Team 2 Economic Trend Sentinel 13_NSI_LT2_EUBD23_pitching slides
Statistics Ireland ESSENTIAL EXPENSE TRACKER 14_NSI_IE_EUBD23_pitching slides
Statistics Lithuania Team 1 lithuania Trends of consumption expenses See the pitching recording
Statistics Austria TrafIO -Transaction Flow Index Observer 16_NSI_AT_EUBD23_pitching slides
ISTAT SEACAT - System for Early Aler on CArd Transactions 17_NSI_IT_EUBD23_pitching slides
INSEE france 4C: Control Credit Card Consumption See the pitching recording
Statistics Hungary Stat Trek DE 19_NSI_HU_EUBD23_pitching slides
Destatis Team 3 TWINNING EUROPE 20_NSI_DE3_EUBD23_pitching slides
MONSTAT Consumption of German resident abroad 21_NSI_ME_EUBD23_pitching slides
Statistics Denmark EUquALL An Economy That Works for All 22_NSI_DK_EUBD23_pitching slides
Czech Statistical Office THE CONSUMPTION BEHAVIOUR OF THE DIFFERENT INCOME GROUPS 23_NSI_CZ_EUBD23_pitching slides
ELSTAT Eureka 24_NSI_EL_EUBD23_pitching slides

 

 

What is a Hackathon?

Data hackathons, also known as data dives, are intense events where teams of data scientists, computer programmers, graphic and interface designers and project managers try to creatively tackle data problems and prototype data analytics products. Hackathons typically last between a day and a week. Some data hackathons are intended simply for educational or social purposes, although in many cases the goal is to create usable data analytics products. Data hackathons tend to have a specific focus, which can include specific data sources, methodologies, technologies and applications but in other cases, there is no restriction on the type of data analytics product being created.

What is the purpose?

The European Big Data Hackathon has five objectives.

  • To solve statistical problems through leveraging algorithms and exploring big data sets as potential sources for official statistics, by engaging with practitioners, developers and data scientists across Europe, to generate ideas and proposals;
  • To build capabilities and identify best-of-class European data scientists, by challenging local developers’ and data scientists’ communities with questions relating to the use of big data for official statistics, by connecting with outstanding participants and teams for possible future collaboration, and by creating networks between statisticians and data scientists for capacity building;
  • To promote and accelerate ‘big data for statistics’ initiatives in Europe, by building prototypes for EU countries to test and integrate in their data development work, and by stirring local communities of scientist-entrepreneurs to work more intensively on big data sourcing, processing and enhancing;
  • To promote partnerships with research community and private sector, by raising awareness about big data developments in Official Statistics in Europe and by initiating, extending and reinforcing collaboration with the private sector and universities.
  • To devise innovative products and tools, including for data visualisation, to stimulate the use of open data and public use files and to engage with new audiences and users.

Programme