Training series and Workshops in Open Science

Bern Data Talks presented by Forschungsdatensupport der UB Bern and Fachstelle Open Science der Berner Fachhochschule.

Topic in the spring semester 2025: AI Meets Open Research Data

Join us for our next Bern Data Talk, focused on the interplay of Artificial Intelligence and Open Research Data (ORD). We will explore the benefits AI can bring to research data management, including improved data analysis and sharing capabilities, while also addressing potential pitfalls and ethical dilemmas. Engage with fellow researchers to share experiences and craft future solutions.

Speakers

• Prof. Dr. Kerstin Denecke (Co-lead Patient-centered Digital Health, Bern University of Applied Sciences, School of Engineering and Computer Science) Navigating the Complexities of Clinical Data for NLP Research - This talk explores the unique challenges of working with clinical data in Natural Language Processing (NLP) research. Unlike other domains where open datasets are common, clinical data remains scarce due to strict privacy regulations. This talk will discuss key hurdles, including data accessibility, legal and ethical considerations, the lengthy annotation process and authorship issues.
• Charles Pierce M.Sc., Nils Tinner (University of Bern, Institute of Geography, Climatology group) AI-Driven Downscaling of ERA5-Land for Urban Temperature Mapping - We present a machine-learning approach to downscale ERA5-Land temperature data using  XGBoost and a convolutional neural network to generate high-resolution urban temperature maps and heat indices. The models are trained for the most part on publicly available datasets from Copernicus services.
• Prof. Dr. Mauricio Reyes (University of Bern, ARTORG Center for Biomedical Engineering Research) Data in Medical Imaging and AI - Quality over quantity / Data is not everything in AI for medical imaging / Sharing experiences

When?

Wednesday, April 2, 2025, 4.15-5.15 pm, followed by an Apéro for in-depth discussions.

Where?

Hochschulstrasse 4, room 105

What are Data Talks?

• Data Talks are informal exchange events about research data and research data management.
• The idea: to give researchers from all disciplines and at all career levels the opportunity to share their experiences, problems, solutions and best practices in dealing with research data.
• The format: once a semester, one topic, three short presentations by researchers, followed by an aperitif that provides space for networking and in-depth discussion with colleagues.
• Would you like to find out more, do you have an idea for a topic or would you like to share your own experiences? Then write to researchdata@unibe.ch. We look forward to hearing from you!

Programs of past Data Talks:

BernDataTalk_Fall_2023.pdf

BernDataTalk_Spring_2024.pdf

BernDataTalk_Fall_2024.pdf

Topic: Protect your Data 

When it comes to research data management, the protection of research data is a crucial aspect. During this one-day event we will combine a short lecture on general concepts of data protection and data management with interactive workshops on strategies and tools to show you different ways for protecting your research data.   

A workshop will only be conducted if at least four participants have registered. 

--------------------------------------------------------------------------------------------------------

When: 18 November, 9:15 - 15:00  

Where: B –102, UniS, Schanzeneckstrasse 1, Bern 

Language: English 

Registration: Link (Deadline: 4 November). Registration is binding; if you are unable to attend, please contact us as soon as possible (researchdata@unibe.ch). 

--------------------------------------------------------------------------------------------------------

Program:  

  9:15 – 10:00: Welcome and introduction 

10:15 – 11:45:

• Workshop 1a: Secure Your Insights: On Running LLMs Locally for Data Privacy 
• Workshop 1b: Secure Data Storage for Humanities Projects 

11:45 – 13:00: Lunch Break 

13:00 – 14:30:

• Workshop 2a: REDCap Survey Setup: Practical and Regulatory Insights 
• Workshop 2b: Anonymisation tools for personal data 

14:45 – 15:00: Wrap up 

--------------------------------------------------------------------------------------------------------

Workshop 1a (10:15 – 11:45):

Secure Your Insights: On Running LLMs Locally for Data Privacy

In response to the critical need for data protection in AI applications, this workshop explores the local deployment of Large Language Models (LLMs). Attendees will engage in practical exercises to configure and interact with LLMs within a secure, offline environment, mitigating risks associated with third-party data handling. This session is designed for researchers and data sceintists who require the power of LLMs for sensitive datasets but must ensure data residency and security.   

Dr. Ing. Federico Grasso Toro, Data Science Data Steward, University of Bern 

Workshop 1b (10:15 – 11:45):

Secure Data Storage for Humanities Projects

Do you have a research project in the Humanities and collect personal data, e.g. in surveys or interviews? This workshop will provide you with basic knowledge and best practices for the secure management of research data. The topics include framework conditions for handling sensitive data (FAIR data principles, Data Protection Act), secure data storage and backup at the University of Bern as well as encryption, access control and secure data deletion. You will also learn how to address SNSF data publication requirements. 

Ursula Loosli M.A., Dr. Gero Schreier, Humanities Data Stewards, University of Bern 

Workshop 2a (13:00 – 14:30):

REDCap Survey Setup: Practical and Regulatory Insights

Are you planning to conduct an online survey for your research project? Or combining a survey along with direct data entry? 

Join us for an interactive workshop to learn how to set up public or personalized surveys in REDCap – a secure data management system with built-in survey functionality. During this hands-on training we will clarify essential Swiss data protection regulations and their impact on survey set-up and implementation. Our workshop is ideal for researchers in all fields and combines legal insight with practical training. 

Basic knowledge of REDCap is an advantage but not required. 

Flurina Jenal M.S, Clinical Data Manager, University of Bern 

Workshop 2b (13:00 – 14:30):

Anonymisation tools for personal data

The handling of personal data in research requires careful data anonymisation with the goal to make any link between research data and the actual individual impossible. In this workshop, we will introduce selected anonymisation tools for different data types. The session offers practical insights into how these tools can support your anonymisation workflows and which limitations you might face when anonymising your data. 

Jennifer Morger M.S., Dr. Christine Krebs, Social and Human Sciences Data Stewards, University of Bern