Data science support

A lot of modern research in the behavioural sciences involves time-intensive and/or multivariate data, think of sensor data, combined with wifi tracking, eye-tracking, and self-reports. Analysing such data sets requires knowledge and expertise in methods that are suitable for such complicated data sets. We can help you extract meaningful insights from your data sets using sophisticated methods. 

We can offer help in using the following advanced data-analytic techniques and help with the interpretation of the results. For example:  

1. Models for time-intensive data 
2. Multilevel models/linear mixed models 
3. Bayesian techniques and/or MCMC methods 
4. Predictive modelling 
5. Time-series modelling 
6. Multidimensional scaling 
7. Item-response theory 
8. Structural equation modelling (SEM) and much more… 

Using abstract, non-representational pictures to illustrate data is a very complex task involving many skills: visual-artistic, empirical-statistical thinking and even logical-mathematical. We see evidence of visualisations every day, from news articles to social media shared infographics. However, most of these visual displays of data often fail to present a clear picture and to enable discovery, insights and analysis. Being able to tell compelling visual stories with your data is very important in todays’ data abundant environment. We can help you develop aesthetically pleasing and effective data visualisations for your projects, from the data structure to publication. 

What we can offer: 

1. Data preparation techniques for visualisation, from dot to story 
2. Data visualisation software and open source codes libraries (e.g., plotly, d3, gephi, dash, ggplot2, etc) 
3. Workshops in basic principles of visualisations, fundamentals of communicating with data, visual and statistical thinking, tools and techniques for building visualisations 
4. Interactive data visualisation systems - escaping the flatland of paper (adding interactivity to your data) 
5. Building interactive tools and dashboards for complex data or analyses (e.g., R shiny apps, dashboards apps, web apps for predictive models) 
6. Visualisation theories for decision making 
7. Fundamentals for displaying quantitative and qualitative information 

We offer a range of support for researchers looking to apply machine learning techniques, from the conception of new lines of research to running large-scale models on a high-performance computing cluster on campus, to building multi-agent generative AI prototypes on local hardware.

CONCEPTION 

We believe it’s important to have an honest discussion about expectations and limitations before we embark on a new data science adventure. We’re excited about machine learning, generative AI and LLMs, and believe that current approaches have much to offer for social science research. At the same time, we’re also aware of the limitations set by the machine learning models, the inherent limits of generative AI, available data, etc. 

While the advancement of artificial intelligence in popular culture often seems to be almost indistinguishable from magic - the reality is that we are a long way from a ‘general’ artificial intelligence. That means that (for now, at least) whether your research questions can be answered through the application of machine learning models depends heavily on the type of question, the quality and amount of training data, and the available resources. 

We’re more than happy to discuss what machine learning applications may be possible for your research, what questions you can ask - and what answers you should expect. 

PREPARING A MODEL 

The success of a machine learning model depends heavily on the care and attention with which it is implemented. Data often needs to be cleaned and reshaped, variables may need to be reshaped or combined to create useful and meaningful features, and model parameters will need to be tuned for best results. 

Each of these steps relies on both technical expertise as well a deep understanding of the subject matter. As such, we believe it is important to have an ongoing conversation between BDSi data science experts and researchers throughout a data science research project. 

Depending on your needs, BDSi can act as a consultant for particularly complex models, or help you think through your entire model from design to implementation. 

RUNNING A MODEL 

BDSi has access to the high-performance computing cluster hosted by the EEMCS (EWI) faculty, allowing us to run large machine learning models on-premise. We also provide local infrastructure to run open-weights LLMs through a chat interface, API, or as a bare-metal platform to develop and test complex AI-based research prototypes. Alternatively, we can also help you run models in a virtual research environment such as the ones offered by LISA. 

INTERPRETING MODEL & RESULTS 

In many cases, being able to interpret the model is as important as the accuracy of the model itself. We can help you get an intuitive understanding of the model and its’ results and create visualizations to communicate with your audience. 

We offer a range of support for researchers looking to apply cutting-edge AI, from exploring new research possibilities and identifying relevant state-of-the-art tools to developing and implementing robust AI solutions.

We have walk-in hours (almost) every monday, from 15:00 to 16:00. During these hours, our data scientists will be available to answer all your questions. We're eager to learn what research you are doing, and explore any opportunities for applying modern data science together.

If we can't help you on the spot, we can set up a consultation later for us to prepare and discuss more in depth, refer you to other experts, or provide you with additional resources.

Consultations are always free. We will aim to provide advice and connect you with other researchers, communities or funding opportunities. Consultations are generally the first step towards collaboration.

To schedule a consultation, please reach out to bdsi@utwente.nl and one of our team members will answer your questions, or contact you for further consultations.

BDSi also collaborates on projects with a high impact for data science at BMS. BDSi Data Science grants are provided several times per year to support projects with a high impact for data science at BMS.

BDSi Data Science grants are meant to:

• kickstart high-risk, high-reward projects with an eye on obtaining further national and international funding at a later date
• support valorization of ongoing BMS research with (interactive) visualization and data-exploration
• explore data-driven approaches to new and existing lines of research by supporting the gathering, storage, and processing of complex datasets

Our hands-on workshops and tutorials offer an interactive learning experience that goes beyond traditional teaching methods. We understand that each researcher has their own style of absorbing information, which is why we design our sessions to cater to diverse skill levels and backgrounds. By creating an environment where you can actively engage with peers and experts, ask burning questions, and receive personalised guidance, we ensure that the knowledge shared is not only relevant but also readily applicable to real-world scenarios in the social sciences.

We help organise events where researchers from various domains come together to discuss shared research interests. These community events provide a platform for networking, sharing ideas, and gaining valuable insights from experienced individuals who have made their mark in the field. We emphasize a diverse range of discussions that are meant to challenge conventions and inspire new research.

Building a strong community is paramount in any field, and data science is no exception. Community building fosters knowledge sharing, collaboration, and growth, creating an environment where individuals can thrive and collectively push the boundaries of their expertise. 

There are numerous existing communities that offer unique perspectives and opportunities for connection. These communities bring together data scientists from diverse backgrounds, experience levels, and specialisations, providing a rich tapestry of ideas and insights. 

In addition to our broad data science community on Teams, we also curate a list of topical communities tailored to specific areas of data science used in the social sciences. Whether you are interested in machine learning, natural language processing, network analysis, or any other niche, there is a community waiting for you. These specialized communities not only offer targeted discussions and resources but also facilitate networking with experts beyond your particular field of application. 

By actively participating in these communities, you can leverage the collective intelligence of the data science community, gain valuable insights, and forge meaningful connections. So, don't miss out on the opportunity to join these vibrant communities. Immerse yourself, contribute your knowledge, learn from others, and together let's shape the future of social data science! 

Part of supporting data science in the long term is building a community of data scientists and researchers interested in data science. Within the community we can share ideas, questions, and expertise with each other. 

The BDSi Data Science Community is free to join for anyone affiliated with the University of Twente or involved in a research project of the University of Twente. 

I have a University of Twente account 

You can instantly join the BDSi Data Science Community team by following these steps: 

1. Open Teams 
2. Go to the "Teams" tab 
3. Make sure you're on the "all teams" page 
4. Click "Join or create team" 
5. Use the code t1p0d1d 

I do NOT have a University of Twente account 

You can apply to join the BDSi Data Science Community team using this Microsoft Teams link.