Automated Methods in Information Visualization
Abstract
Visualization and Machine Learning have related goals in terms of helping analysts to understand characteristic aspects of data. While visualization aims at involving the user through interactive depictions of data, machine learning is generally represented by automatic methods that yield optimal results with respect to certain initially specified tasks. Not at the least within the research direction of visual analytics it seems promising to think about opportunities to integrate both methodologies in order to exploit the strengths of both sides. Up to now, examples of integration very often encompass the visualization of results from automatic methods as well as attempts to make originally automated methods partially interactive. A vision for the future would be to integrate interactive and automatic methods in order to solve problems. A possible realization could be an iterative process where the one or other approach is chosen on demand at each step.
H. Hauser, Automated Methods in Information Visualization, 2012.
[BibTeX]
Visualization and Machine Learning have related goals in terms of helping analysts to understand characteristic aspects of data. While visualization aims at involving the user through interactive depictions of data, machine learning is generally represented by automatic methods that yield optimal results with respect to certain initially specified tasks. Not at the least within the research direction of visual analytics it seems promising to think about opportunities to integrate both methodologies in order to exploit the strengths of both sides. Up to now, examples of integration very often encompass the visualization of results from automatic methods as well as attempts to make originally automated methods partially interactive. A vision for the future would be to integrate interactive and automatic methods in order to solve problems. A possible realization could be an iterative process where the one or other approach is chosen on demand at each step.
@MISC {Hauser12Dagstuhl,
author = "Helwig Hauser",
title = "Automated Methods in Information Visualization",
howpublished = "Invited talk at the Dagstuhl seminar 12081",
month = "February",
year = "2012",
abstract = "Visualization and Machine Learning have related goals in terms of helping analysts to understand characteristic aspects of data. While visualization aims at involving the user through interactive depictions of data, machine learning is generally represented by automatic methods that yield optimal results with respect to certain initially specified tasks. Not at the least within the research direction of visual analytics it seems promising to think about opportunities to integrate both methodologies in order to exploit the strengths of both sides. Up to now, examples of integration very often encompass the visualization of results from automatic methods as well as attempts to make originally automated methods partially interactive. A vision for the future would be to integrate interactive and automatic methods in order to solve problems. A possible realization could be an iterative process where the one or other approach is chosen on demand at each step. ",
images = "images/Hauser12Dagstuhl.png",
thumbnails = "images/Hauser12Dagstuhl_thumb.png",
location = "Wadern, Germany",
url = "//www.dagstuhl.de/en/program/calendar/semhp/?semnr=12081",
pres = "pdfs/Hauser12Dagstuhl-pres.pdf"
}