ADIB: Adaptive Information Support
Project Topic
Based on an integrated task management system,
ADIB aims to develop and test a tool which adaptively and pro-actively provides information and know-how for knowledge intensive tasks.
Project Description
ADIB's goals are
- Developing methods for proactive information support which
- capture the user's task context by continuous user observation
- adapt the information support to the task context dynamically
- consider explicit user feedback to adapt the user model and the task context
- Identification, systematization, and unification of different task concepts in knowledge work processes
- Similarity-based reuse of everyday work practices with the help of
- a case-based retrieval of hierarchically structured task descriptions
- techniques of layout-analysis to determine structural document similarity and extraction of information from the document
- statistical language models for reliable information extraction
To reach these goals,
ADIB is based on
- desktop user observation done in "EPOS"
- DFKI Open Source search tool for dynamic queries DynaQ
- similarity measures for object-oriented case representations and learning of knowledge-intensive similarity-measures for case-based reasoning
- geometric and logical layout analysis using layout analysis by exploring the space of segmentation parameters and by using layout analysis by example
Project Members
Project Chair
Participating Research Groups
- Research Department "Knowledge Management", DFKI
- Knowledge-Based Systems Group, Department of Computer Science
- Research Department "Image Understanding and Pattern Recognition", DFKI
- Image Understanding and Pattern Recognition Group, Department of Computer Science
Scientific Personnel
- Dr. Heiko Maus (Research Department "Knowledge Management", DFKI)
- Dr. Armin Stahl (Research Department "Image Understanding and Pattern Recognition", DFKI)
- Joost van Beusekom (Image Understanding and Pattern Recognition Group, Department of Computer Science)
- Kinga Schumacher (Research Department "Knowledge Management", DFKI)
Former member
- Dr. Harald Holz (Research Department "Knowledge Management", DFKI)
External Cooperation
- EU IP Projekt NEPOMUK
- BMBF Projekte Mymory, IPeT;
- Rhineland-Palatinate funded project DynaQ
- Prof. Hiroshi Tsuji, Osaka Prefecture University, Japan
- student Shingo Kubo at DFKI from July 2007 to January 2008, topic Trend Recognition in the Semantic Desktop
- Mark Siebert, SBS München
Project Events and Achievements
- Project start date: December 1st, 2005
- Project end date: December 31st, 2007
Project Publications
Kinga Schumacher, Michael Sintek, Leo Sauermann. In:
Proceedings of the 5th European Semantic Web Conference (ESWC 08). LNCS, Volume 5021, Springer, P. 569--583, 2008
Kinga Schumacher. In:
12th International Conference on Applications of Natural Language to Information Systems (NLDB), CNAM, Paris, France. LNCS, Volume 4592, Springer, June, 2007
Heiko Maus. TU Kaiserslautern, dissertation.de - Verlag im Internet GmbH, ISBN 978-3-86624-267-8, 2007 (in german)
Heiko Maus, Andreas Dengel. In: Koichi Kise and David S. Doermann ed.,
Proceedings of Second International Workshop on Camera-Based Document Analysis and ecognition (CBDAR2007). 2007
Armin Stahl. In:
Proceedings of the 8th European Conference on Case-Based Reasoning (ECCBR 2006). Springer, 2006
Joost van Beusekom, Daniel Keysers, Faisal Shafait, Thomas M. Breuel. In:
International Conference on Document Image Analysis for Libraries (DIAL 2006). IEEE, 2006
Mark Siebert, Pierre Smits, Leo Sauermann, Andreas Dengel. In:
Proceedings EDM, IEEE Int. Workshop on the Electronic Document Management in an Enterprise Computing Environment, Hong Kong, China. IEEE Computer Society, 2006
Mark Siebert, Pierre Smits, Leo Sauermann, Andreas Dengel. In: Ulrich Reimer and Dimitris Karagiannis ed.,
PAKM 2006. LNAI, Volume 4333, Springer, P. 279--290, 2006
Presentations
Tools
- myCBR is a case based reasoning tool that can both be used as standalone application and as a Protégé plugin. myCBR is open source, developed under the GPL license. More details at the myCBR project homepage