Informatics, TU Vienna

Exploratory Search in Time-Oriented Data

In his thesis, Jürgen Bernard addressed analytical challenges related to the increasing amounts of time-oriented data.

Abstract

In his thesis, Jürgen Bernard addressed analytical challenges related to the increasing amounts of time-oriented data. Particularly, application fields related to data-driven research are confronted with analytical problems such as exploring data or asking meaningful questions about large and previously unknown data sets. He showed how "Exploratory Search" can help researchers to gain an overview of large and unknown data sets, and subsequently formulate meaningful queries to retrieve interesting subsets of data. All of his solutions have in common that visual-interactive interfaces are presented as the means to analyze time-oriented data in a meaningful way. At a glance, He presented concepts, guidelines, techniques, and applications that can help future data analysis approaches to support data-driven research. In two design studies, Jürgen Bernard put his contributions into practice and showed how researchers can benefit from visual-interactive Exploratory Search Systems.

In his current line of research as a Postdoc at TU Darmstadt, he focuses on the challenge of segmenting and labeling multivariate time-oriented data. Meaningful segments of multivariate time-oriented data are highly beneficial for many downstream analyses and can be applied in a variety of application areas. However, the sheer amount of algorithmic models that can be used to process, segment, and label multivariate time -oriented data often exceeds analysts' abilities to create most meaning results. In addition, the parameterization of respective models raises the number of possible solutions to higher power. Finally, different types uncertainty of data and models requires a more in-depth examination. More information here.

Biography

Dr. Jürgen Bernard is working as a post-doctorate (postdoc) at TU Darmstadt in the Interactive Graphics Systems Group (GRIS). His research combines Information Visualization, Visual Analytics, Exploratory Search, Data Mining, and Machine Learning. His PhD thesis was about "Exploratory Search in Time-Oriented Primary Data". He presented novel concepts, guidelines, techniques, and systems to support domain experts in data-centered research in the area of big data. For his thesis, he was awarded with the Hugo-Geiger Preis 2016 for excellent junior researchers provided by the Fraunhofer Society.

Note

This talk is organized by the Information and Software Engineering Group at the Institute of Software Technology and Interactive Systems.