Visualization plays an important role throughout the entire scientific process because it forms the link between the computer and the scientist. At the eScience Center we manage projects that range from raw scientific data visualization to infographics and visual storytelling. Many of our current projects combine data visualization with literate programming in online interactive notebook environments.
An important part of quantitative research is the estimation of the uncertainty of the results and exploring the sources of error with statistics. At the eScience Center we apply statistics in domains ranging from psychology to astronomy. We investigate specific sampling and analysis techniques, setting up good surveys and new statistical methods such as multilevel modeling.
Natural language is one of the main ways in which people communicate. In the past decades, big steps ahead have been made to teach computers how to read in order to, for example, automate or assist literature research. At the eScience Center we work on NLP research ranging from social media analysis to the digitalization of archives. Topics we work on include topic modeling, sentiment analysis, machine translation and optical character recognition.
We currently live in the boom time of machine learning. Every single field of research is being accelerated by the application of machine learning, sometimes to automate tasks, sometimes to find patterns invisible to a human researcher. At the eScience Center we work on both supervised and unsupervised machine learning, applied to tasks ranging from the prediction of chemical properties to natural language understanding. We work on applications of deep learning, explainable AI and time series classification.
At the eScience Center we use data mining to find patterns in data for discovering new hypotheses or answer existing questions. Examples are text mining, finding new patterns in climate data to predict abrubt shifts and monitoring the workings of our own organization.
At the eScience Center we work on computer vision to recognize objects, persons and behavior in still images and video. Automating visual analysis enables researchers to work with image and video collection of a much larger size than they would be able to analyze personally. The research projects we work on include the analysis of satellite imagery for the classification of land use, and detecting when plants start to grow or shed leaves and when they bloom, the analysis of medical scans to detect anomalies or the the analysis of scans for optical character recognition for automated digitalization of old documents.