The document summarizes the explanation capabilities of the open source case-based reasoning tool myCBR. It discusses how myCBR provides conceptual explanations to build links between unknown and known concepts, and action explanations to describe the similarity calculation process and retrieval results. The document also presents examples of different types of questions myCBR can answer through explanations during case retrieval and system modeling and maintenance.
Explanation Capabilities of the Open Source Case-Based Reasoning Tool myCBR
1. CAMBRIDGE, UK, 9 DEC 2008
Explanation Capabilities of the Open Source
Case-Based Reasoning Tool myCBR
Thomas Roth-Berghofer
Senior researcher, trb@dfki.de
German Research Centre for Artificial Intelligence DFKI GmbH
Samstag, 18. Juli 2009
7. What are
explanations?
Explanations
are answers to
questions.
Samstag, 18. Juli 2009
8. • Plug-in for popular ontology editor Protégé (Version 3.x)
• Extensions provided by myCBR (Version 2.6.1):
• Easy import of raw data (csv-files)
• GUIs for modelling knowledge-intensive similarity measures
• Similarity-based retrieval functionality
• Export of domain model and similarity measures in XML format
• Stand-alone retrieval engine for separate use or
integration with other systems
• Built-in explanation capabilities
MORE INFORMATION AND DOWNLOAD: HTTP://MYCBR-PROJECT.NET
Samstag, 18. Juli 2009
15. Global similarity measure
• Feature match:
local similarity
measures
• Class
similarity:
aggregate local
similarities
• Weighted sum
• Maximum
• Minimum
• Euclidian distance
Samstag, 18. Juli 2009
16. Global similarity measure
• Feature match:
local similarity
measures
• Class
similarity:
aggregate local
similarities
• Weighted sum
• Maximum
• Minimum
• Euclidian distance
Samstag, 18. Juli 2009
25. Questions
• about domain concepts
• about retrieval results
• during modelling and maintenance
Samstag, 18. Juli 2009
26. Question about
domain concepts
• What is meant by this concept?
Samstag, 18. Juli 2009
27. Conceptual Explanations
• The goal of conceptual explanations is to build links
between unknown and known concepts.
• Variations:
• Definition: “What is a bicycle?” – “A bicycle is a land vehicle
with two wheels in line. Bicycles are a form of human powered
vehicle.”
• Theoretical proposition: “What is force?” – “Force is Mass
times Acceleration.”
• Prototypical usage of individual things or actions: “What is
a bicycle?” – “The thing, that man over there just crashed
with.”
• Functional mapping: “What is a bicycle?” – “A bicycle serves
as a means of transport.”
Samstag, 18. Juli 2009
35. Questions about
retrieval results
• How did the system come to the similarity
assessment of a particular case?
Samstag, 18. Juli 2009
36. Questions about
retrieval results
• How did the system come to the similarity Action
assessment of a particular case? explanations
Samstag, 18. Juli 2009
37. Questions about
retrieval results
• How did the system come to the similarity Action
assessment of a particular case? explanations
• Which are the most similar aspects of a
case? Which are the least?
Samstag, 18. Juli 2009
38. Action explanations
• Action explanations explain the activities of the
respective system.
• Action explanations:
“Why was this seat post selected?” –
“For the given price, only one other seat
post was available. But this was too
short.
• Negative action explanations:
“Why was no carrier chosen?” –
“A carrier is only available for
touring bikes. The user did not
choose a touring bike.”
Samstag, 18. Juli 2009
45. • Conceptual Explanations provide
links to known concepts.
• Action explanations provide
information about the similarity
calculation process.
Samstag, 18. Juli 2009
46. Most / least important
aspects of a case
Samstag, 18. Juli 2009
47. Most / least important
aspects of a case
Samstag, 18. Juli 2009
50. Questions during modelling
and maintenance
• Are some problem types underrepresented in the
case base?
• Is there an imbalance of cases in the case base?
Samstag, 18. Juli 2009
60. Take home messages
• Explanations are important and
need to be considered while
developing a software system.
Samstag, 18. Juli 2009
61. Take home messages
• Explanations are important and Explainer
need to be considered while User
developing a software system.
Originator
• Explanations are part of
communication scenario
Samstag, 18. Juli 2009
62. Take home messages
• Explanations are important and Explainer
need to be considered while User
developing a software system.
Originator
• Explanations are part of
communication scenario
• Conceptual explanations support
domain understanding.
Samstag, 18. Juli 2009
63. Take home messages
• Explanations are important and Explainer
need to be considered while User
developing a software system.
Originator
• Explanations are part of
communication scenario
• Conceptual explanations support
domain understanding.
• Action explanations counteract
encapsulation and information
hiding.
Samstag, 18. Juli 2009
64. Take home messages
• Explanations are important and Explainer
need to be considered while User
developing a software system.
Originator
• Explanations are part of
communication scenario
• Conceptual explanations support
domain understanding.
• Action explanations counteract
encapsulation and information
hiding.
• Explanation manager provides
access to conceptual and action
explanations.
Samstag, 18. Juli 2009
65. Thank you!
CAMBRIDGE, UK, 9 DEC 2008
Explanation Capabilities of the Open Source
Case-Based Reasoning Tool myCBR
Thomas Roth-Berghofer
Senior researcher, trb@dfki.de
German Research Centre for Artificial Intelligence DFKI GmbH
Samstag, 18. Juli 2009
66. Invitation to participate
• ExaCt mailing list:
http://groups.yahoo.com/group/explanation-research/
• ExaCt 2009 @ IJCAI 2009
http://exact2009.workshop.hm
Samstag, 18. Juli 2009
67. Publications of ExaCt
research group
2008 2007 (contd.)
[Bahls, 2008] Bahls, D. (2008). Explanation support for the case-based [Roth-Berghofer et al., 2007] Roth-Berghofer, T. R., Schulz, S., and Leake,
reasoning tool myCBR. Project thesis, University of Kaiserslautern. D. B., editors (2007). Proceedings of the AAAI-07 workshop on
[Forcher et al., 2008] Forcher, B., Adrian, B., and Roth-Berghofer, T. (2008). Explanation-aware Computing ExaCt2007. AAAI Press. Technical Report
Explanations in the information extraction system iDocument. Künstliche WS-07-06.
Intelligenz, 22(2). 2006
[Mittag, 2008] Mittag, F. (2008). ReduxExp: A justification-based [Richter et al., 2006] Richter, M. M., Roth-Berghofer, T., and Schulz, S.,
explanation-support server. Project thesis, University of Kaiserslautern. editors (2006). Explanation-aware Computing, volume 25. SAP - Slovak
[Roth-Berghofer and Bahls, 2008] Roth-Berghofer, T. R. and Bahls, D. Academic Press Ltd., Bratislava.
(2008). Code tagging and retrieval with myCBR. In Petridis, M., Coenen, F., 2005
and Bramer, M., editors, Research and Development in Intelligent Systems [Roth-Berghofer et al., 2005a] Roth-Berghofer, T., Cassens, J., and Sørmo,
XXV, London, UK. Springer Verlag. F. (2005a). Goals and kinds of explanations in case-based reasoning. In
[Roth-Berghofer and Mittag, 2008] Roth-Berghofer, T. R. and Mittag, F. Althoff, K.-D., Dengel, A., Bergmann, R., Nick, M., and Roth-Berghofer, T.,
(2008). ReduxExp: A justification-based explanation-support server. editors, WM 2005: Professional Knowledge Management, pages 264–268,
Proceedings of AI-2008. the twenty-eighth SGAI international conference Kaiserslautern, Germany. DFKI GmbH.
on artificial intelligence. In Petridis, M., Coenen, F., and Bramer, M., editors,
[Roth-Berghofer and Cassens, 2005] Roth-Berghofer, T. R. and Cassens, J.
Research and Development in Intelligent Systems XXV, London, UK.
(2005). Mapping goals and kinds of explanations to the knowledge
Springer Verlag.
containers of case-based reasoning systems. In Muñoz-Avila, H. and Ricci,
[Roth-Berghofer and Richter, 2008a] Roth-Berghofer, T. R. and Richter, M. F., editors, Case-Based Reasoning Research and Developmen, pages 451–
M., editors (2008a). Künstliche Intelligenz—Topic: Explanation, volume 22, 464, Heidelberg. Springer Verlag.
Bremen. BöttcherIT Verlag.
[Roth-Berghofer et al., 2005b] Roth-Berghofer, T. R., Schulz, S., and
[Roth-Berghofer and Richter, 2008b] Roth-Berghofer, T. R. and Richter, M. Woody, A., editors (2005b). Proceedings of the AAAI Fal l Symposium on
M. (2008b). On explanation. Künstliche Intelligenz, 22(2):5–7. Explanation-aware Computing ExaCt2005. AAAI Press. Technical Report
2007 FS-05-04.
[Bahls and Roth-Berghofer, 2007] Bahls, D. and Roth-Berghofer, T. (2007). 2004
Explanation support for the case-based reasoning tool myCBR. In [Roth-Berghofer, 2004] Roth-Berghofer, T. R. (2004). Explanations and
Proceedings of the Twenty-Second AAAI Conference on Artificial Case-Based Reasoning: Foundational issues. In Funk, P. and González-
Intelligence. July 22–26, 2007, Vancouver, British Columbia, Canada., Calero, P. A., editors, Advances in Case-Based Reasoning, pages 389–403.
pages 1844–1845. The AAAI Press, Menlo Park, California. Springer-Verlag.
[Eppert, 2007] Eppert, M. (2007). Generating provenance explanations for
the gnowsis rebirth machine - a first pass. Pro ject thesis, University of
Kaiserslautern.
Samstag, 18. Juli 2009