The ODISSEI Hub develops a community and supports the usage and usability of the infrastructure through an educational program, community events and as analytic support.

It is a work stream in the ODISSEI Roadmap project. The other work streams are the Data Facility, Observatory, and Laboratory.

The Hub consists of five closely interrelated tasks:

  1. Community management
  2. Educational programme
  3. Social data science team
  4. Benchmarking
  5. Grants for computational social science

4.1 Community Management

Community management is aimed at building and maintaining a sustainable ODISSEI collectivity. It is led by the Coordination Team based at EUR who have several responsibilities within ODISSEI to ensure smooth operations across the diverse range of activities. The first subtask will be the general administration of the project including financial and technical reporting, the administration of calls and the monitoring of progress. Being a community infrastructure, it is vital that ODISSEI actively monitors researchers’ needs as input for future directions and communicates what ODISSEI offers. Community Officers will engage in various community development activities through regular face-to-face meetings as well as mailings, the website, social media, and the aforementioned ODISSEI annual conference. Community Officers are the first point of contact for researchers from member organisations turning to ODISSEI. Their job is to help researchers find their way in the infrastructure. Finally, Data Stewards will help researchers create FAIR data and execute ethical practices. Experts from DANS will develop and maintain the user agreement, which focus on data management, data sharing, research ethics, and the principles and practice of FAIR data. ODISSEI Data Stewards will provide support to researchers to implement the user agreement, in collaboration with the experts at DANS.

Project team Community Management: Lucas van der Meer (ODISSEI Coordination Team – Task leader).

4.2 Educational programme

The ODISSEI Hub’s External Relations Officer will operate an extensive educational programme to facilitate the use of the infrastructure and the uptake of computational and innovative methods in the social sciences. First, there will be an annual ODISSEI Conference where researchers can showcase their projects and engage in broad and open discussion about developments in the infrastructure. This annual event will be hosted by a different member organisation every year and will be coordinated by EUR. Second, there will be a series of workshops aimed at promoting the usage of the infrastructure and the methods and skills needed to do so. These workshops will be coordinated by the EUR but will be hosted by various member organisations. There will be approximately six workshops per year. The program of workshops includes but is not limited to: The ODISSEI Data Facility users Bootcamp; An introduction to Statistics Netherlands microdata access; An introduction to the LISS Panel; Ethics and GDPR in ODISSEI; Machine learning for social science;. Third, there will be regular communications with the educational officers at member organisations to synchronise and coordinate existing educational programs on computational social sciences.

Project team Educational programme: Lucas van der Meer (ODISSEI Coordination Team – Task leader).

4.3 Social data science team

The Social Data Science Team will bridge the gap between applied social scientists and the data, computing, and analytical infrastructure provided by ODISSEI. They will actively seek out short-term collaborative projects with social scientists across ODISSEI member organisations to uniquely leverage ODISSEI’s infrastructure to answer novel substantive questions and to build expertise in computational social science. These projects will result not only in ‘traditional’ publications, but also in open source reusable code that can be used for teaching purposes or as a starting point for other projects.

(Project team Social data science team: Daniel Oberski (Utrecht University – Task leader), Erik-Jan van Kesteren (Utrecht University).

Questions regarding Social data science team? Contact Kasia Karpinska (ODISSEI Coordination Team).

4.4 Grants for computational social science

The ODISSEI Hub will provide grants for computational social science to support the use of the data infrastructure developed across the other three Work streams. Following a successful practice employed at the NLeSC, these grants will provide hours from eScience Research Engineers employed at the NLeSC to collaborate with social scientists to enhance their research, by exploiting digital technology. The grants will be made available via an annual open call which will be assessed by experts in the field of computational social science.

Project team Grants for computational social science: Frank Seinstra (Netherlands eScience Center – Task leader).

Questions regarding Grants for computational social science? Contact Kasia Karpinska (ODISSEI Coordination Team).

4.5 Benchmarking

The ODISSEI Hub will pilot workshops and hackathons on computational benchmarking in the social sciences. In computational sciences, researchers often have to choose between different computational approaches. Benchmarking uses standard datasets and parameters to systematically compare computational approaches and establish best practices in the community. Since translating research problems into benchmark studies is a novel concept in social sciences, workshops will be organised on their design and to build consensus on the set-up for a specific task. This will eventually allow ODISSEI to monitor the state-of-the-art algorithms for a specific task on the EYRA (Enhance Your Research Alliance) Benchmark Platform developed by NLeSC and SURFsara and will be a community effort. Donoho (2017) ascribed a large part of the astounding recent successes in machine learning to the application of the ‘common task framework’ (CTF), in which (1) a training dataset derived from ODISSEI infrastructure is publicly shared, (2) teams of experts compete for a common task pertaining to prediction on the training data, and (3) an impartial referee reports a performance score for each team based on held-out testing data. Fields in which this competition framework has been systematically applied, such as automated driving, machine translation, and medical imaging, have reported impressive and objectively quantifiable progress (Donoho, 2017).

Project team Benchmarking: Adriënne Mendrik (Netherlands eScience Center – Task leader).

Questions regarding Benchmarking? Contact Kasia Karpinska (ODISSEI Coordination Team).