Data Science

Research for tomorrow's smart systems

Today, data stems from sensors and electronic processes, from social media and the web, and it is generated by smartphones, cars, and machines everywhere. Data sets are big yet incomplete, noisy yet rich in information, rapidly growing and complex yet in need of understanding. We investigate machine learning and data mining technologies that take up the challenges of real data and practical information needs.

True to Fraunhofer’s mission, we research and develop data analytics technologies motivated both from our customers' needs and from future trends in society and the industry. We take up the challenges of ubiquitous digitalization and digital lifestyles, and fuel the smart systems of tomorrow. 

 

Research topics

Big Data Architecture and Analytics

We solve complex analysis tasks with big data sets under realtime conditions.

Machine Learning

Where one-size-fits-all solutions can no longer cope with the challenges of big data, we develop smarter and faster approach to machine learning.

Interactive and Visual Analytics

Often, a picture can say more than 1 000 words and a graphic says more than long rows of data. Smart systems understand their users and produce helpful vsualizations.

Text Analytics

Texts hide knowledge that waits to be discovered. We develop algorithms which extract semantics from different types of unstructured data.

Deep Learning

Neural networks  are trainiered with huge sets of data.  Fraunhofer IAIS is a pioneer in developing deep learning approaches  for German companies.

Natural Language Question Answering

Queston-answering systems are becoming an inspiring model for future search engines. We investigate systems based on increasing assets of linked data.

Application areas

Our applied research is driven by our customers' needs, by evolving trends in society, and by upcoming challenges in industry. Application areas include:

  • Finance: supervised and unsupervised techniques for fraud detection
  • Healthcare analytics, such as the analysis of networks, clinical decision support, or information extraction from medical texts
  • Industry 4.0: analysis of high-frequency sensor data and collaborative methods for a data-driven support of engineers
  • Advanced analytics in telecommunications, logistics, and the automotive industry