Detecting real user tasks by training on laboratory contextual attention metadata

Andreas S. Rath; Didier Devaurs; Stefanie N. Lindstaedt

Detecting real user tasks by training on laboratory contextual attention metadata

Andreas S. Rath^*, Didier Devaurs, Stefanie N. Lindstaedt

^*Corresponding author for this work

Computer And Information Sciences

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution book

2 Citations (Scopus)

Abstract

Detecting the current task of a user is essential for providing her with contextualized and personalized support, and using Contextual Attention Metadata (CAM) can help doing so. Some recent approaches propose to perform automatic user task detection by means of task classifiers using such metadata. In this paper, we show that good results can be achieved by training such classifiers offline on CAM gathered in laboratory settings. We also isolate a combination of metadata features that present a significantly better discriminative power than classical ones.

Original language	English
Title of host publication	INFORMATIK 2009 - Im Focus das Leben, Beitrage der 39. Jahrestagung der Gesellschaft fur Informatik e.V. (GI)
Place of Publication	Bonn, Germany
Pages	1645-1653
Number of pages	9
Publication status	Published - 28 Sept 2009
Event	39th Jahrestagung der Gesellschaft fur Informatik e.V. (GI): Im Focus das Leben, INFORMATIK 2009 39th Annual Meeting of the German Informatics Society (GI): Focus on Life, INFORMATIK 2009 - Lubeck, Germany Duration: 28 Sept 2009 → 2 Oct 2009

Conference

Conference	39th Jahrestagung der Gesellschaft fur Informatik e.V. (GI): Im Focus das Leben, INFORMATIK 2009 39th Annual Meeting of the German Informatics Society (GI): Focus on Life, INFORMATIK 2009
Country/Territory	Germany
City	Lubeck
Period	28/09/09 → 2/10/09

Keywords

discriminative power
task detection

Cite this

@inproceedings{e818bc603ad24302ad81911d77269ac2,

title = "Detecting real user tasks by training on laboratory contextual attention metadata",

abstract = "Detecting the current task of a user is essential for providing her with contextualized and personalized support, and using Contextual Attention Metadata (CAM) can help doing so. Some recent approaches propose to perform automatic user task detection by means of task classifiers using such metadata. In this paper, we show that good results can be achieved by training such classifiers offline on CAM gathered in laboratory settings. We also isolate a combination of metadata features that present a significantly better discriminative power than classical ones.",

keywords = "discriminative power, task detection",

author = "Rath, {Andreas S.} and Didier Devaurs and Lindstaedt, {Stefanie N.}",

year = "2009",

month = sep,

day = "28",

language = "English",

isbn = "9783885792482",

pages = "1645--1653",

booktitle = "INFORMATIK 2009 - Im Focus das Leben, Beitrage der 39. Jahrestagung der Gesellschaft fur Informatik e.V. (GI)",

note = "39th Jahrestagung der Gesellschaft fur Informatik e.V. (GI): Im Focus das Leben, INFORMATIK 2009 39th Annual Meeting of the German Informatics Society (GI): Focus on Life, INFORMATIK 2009 ; Conference date: 28-09-2009 Through 02-10-2009",

}

Rath, AS, Devaurs, D & Lindstaedt, SN 2009, Detecting real user tasks by training on laboratory contextual attention metadata. in INFORMATIK 2009 - Im Focus das Leben, Beitrage der 39. Jahrestagung der Gesellschaft fur Informatik e.V. (GI). Bonn, Germany, pp. 1645-1653, 39th Jahrestagung der Gesellschaft fur Informatik e.V. (GI): Im Focus das Leben, INFORMATIK 2009 39th Annual Meeting of the German Informatics Society (GI): Focus on Life, INFORMATIK 2009, Lubeck, Germany, 28/09/09.

Detecting real user tasks by training on laboratory contextual attention metadata. / Rath, Andreas S.; Devaurs, Didier; Lindstaedt, Stefanie N.
INFORMATIK 2009 - Im Focus das Leben, Beitrage der 39. Jahrestagung der Gesellschaft fur Informatik e.V. (GI). Bonn, Germany, 2009. p. 1645-1653.

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution book

TY - GEN

T1 - Detecting real user tasks by training on laboratory contextual attention metadata

AU - Rath, Andreas S.

AU - Devaurs, Didier

AU - Lindstaedt, Stefanie N.

PY - 2009/9/28

Y1 - 2009/9/28

N2 - Detecting the current task of a user is essential for providing her with contextualized and personalized support, and using Contextual Attention Metadata (CAM) can help doing so. Some recent approaches propose to perform automatic user task detection by means of task classifiers using such metadata. In this paper, we show that good results can be achieved by training such classifiers offline on CAM gathered in laboratory settings. We also isolate a combination of metadata features that present a significantly better discriminative power than classical ones.

AB - Detecting the current task of a user is essential for providing her with contextualized and personalized support, and using Contextual Attention Metadata (CAM) can help doing so. Some recent approaches propose to perform automatic user task detection by means of task classifiers using such metadata. In this paper, we show that good results can be achieved by training such classifiers offline on CAM gathered in laboratory settings. We also isolate a combination of metadata features that present a significantly better discriminative power than classical ones.

KW - discriminative power

KW - task detection

M3 - Conference contribution book

AN - SCOPUS:84874336919

SN - 9783885792482

SP - 1645

EP - 1653

BT - INFORMATIK 2009 - Im Focus das Leben, Beitrage der 39. Jahrestagung der Gesellschaft fur Informatik e.V. (GI)

CY - Bonn, Germany

T2 - 39th Jahrestagung der Gesellschaft fur Informatik e.V. (GI): Im Focus das Leben, INFORMATIK 2009 39th Annual Meeting of the German Informatics Society (GI): Focus on Life, INFORMATIK 2009

Y2 - 28 September 2009 through 2 October 2009

ER -

Detecting real user tasks by training on laboratory contextual attention metadata

Abstract

Conference

Keywords

Fingerprint

Cite this