Jožef Stefan Institute (JSI), founded in 1949 and employing around 1,000 staff, is the leading Slovenian institution for research in natural sciences and technology. It has a long tradition of developing partnerships with the industry, as well as with various institutions across the globe, as evidenced by the current participation in over 300 international projects.
The Department of Intelligent Systems (DIS) conducts research in several areas of artificial intelligence. Our foremost expertise is machine learning, where we specialize in adapting classifiers to particular users and situations by means of semi-supervised learning, using heterogeneous context information to improve the performance of machine-learning models, and combining machine learning with expert knowledge to increase the transparency and robustness of such models. We also have expertise in decision support, optimization, multi-agent systems, speech and language technologies. Our most important application domains are e-health and ambient intelligence. We analyse and interpret human behaviour using sensor and other data. We have also developed a number of methods for activity recognition, human energy expenditure estimation and similar. We also develop methods to assess health using biomedical signals, electronic health records and user feedback, a prime example of which is our diagnostic method in the finals of the 10-million-USD XPrize Tricorder competition.
While DIS pursues basic research in artificial intelligence, our main strength is probably the application of intelligent computer methods to various real-life problems. We also have extensive experience with international collaboration, having participated in many FP6, FP7, H2020 and other international projects.
In CrowdHEALTH, JSI uses its extensive experience with the analysis of sensor and other data to interpret the types of source data that are not immediately usable, particularly those coming from mobile devices. JSI then contributes to the analysis of the information in the HHRs and extraction of knowledge. Finally, JSI works on the visualization of this information and knowledge, and also works closely with ULJ on the use case aimed at reduction of obesity and improvement of physical activity and fitness of children and youth.