University of Konstanz
Graduiertenkolleg / PhD Program
Computer and Information Science

Graduation Talks

title

Visual Analytics in Online Information Streams

speaker

Milos Krstajic, University Konstanz
Konstanz, Germany

date & place

Wednesday, 19.01.2011, 16:15 h
Room C 252

abstract

A vast amount of news is generated every day by an increasing number of sources and the analysis of these data requires automated and interactive visual meth- ods that will facilitate processing of the acquired information for the analysts in different domains. News data items in online information streams are character- ized by textual and temporal attributes, and analyzing the relationships between these attributes and across data presents one of the fundamental problems in text data mining. News articles come from various different sources, in differ- ent amount and frequency, and the text of the articles is often copied, modified and filtered across the sources. The topics that are appearing in the news are characterized by high temporary changes, hierarchical structure and inter-topic relationships. Analyzing the dynamics, i.e. the change in amount, frequency and content of this type of data, in a streaming environment defines several challenges in our research.