Dr. Dimitrios Tzovaras
Director, Researcher A
Information Technologies Institute/ Centre for Research and Technology Hellas
Visual Analytics Technologies for the Efficient Processing and Analysis of Big Data
Nowadays, there is a vast amount of data produced on daily basis from a great variety of sources, ranging from personal multimedia sensors to big health related datasets and network events. The issue of being able to analyse them in reasonable time, to comprehensively present them and generally to process them in an efficient manner regardless their eventual application is most commonly referred to as the “Big Data” issue and forms a great challenge to the research society. As a consequence, any technology, that claims to overcome the information overload problem, has to provide answers for the following questions (a) Who or what defines the relevance of information” for a given task, (b) How can appropriate procedures in a complex decision making process be identified (c) How can the information be presented in a decision- or task-oriented way (d) Which types of interaction can facilitate problem solving and decision making.
The goal of visual analytics research is to turn the information overload into an opportunity by enabling decision-makers to examine this massive amount of information to make effective decisions. Visual analytics enables people to find hidden relations and to turn the data into useful and defensible knowledge. In this respect, the current presentation, held by Dr. Dimitrios Tzovaras, aims to revisit the latest achievements of the Visual Analytics, Virtual & Augmented Reality Laboratory of CERTH/ITI in the field of Big Data Analysis. In particular, by relying on advanced visual analytics, by extending Graph-based approaches and by utilizing beyond state-of-the-art methods for significant dimensionality reduction of the solution domain (i.e. pareto-front), efficient solutions have been proposed and evaluated for several Big Data related applications. Such applications included but are not limited to: internet and telecommunication networks security, multimedia search engines, road traffic management and prediction, as well as pattern detection in large health databases (i.e. DNA sequences).
The keynote speech will be hosted at Sunday 21 Sep 2014 15.30 at the Room A.