2004 Fiscal Year Final Research Report Summary
Development of Cyber ADR system
| Project/Area Number |
14580445
|
|---|
| Research Category |
Grant-in-Aid for Scientific Research (C)
|
| Allocation Type | Single-year Grants |
|---|
| Section | 一般 |
|---|
| Research Field |
情報システム学(含情報図書館学)
|
|---|
| Research Institution | Tokyo Institute of Technology |
|---|
Principal Investigator |
NITTA Katsumi Tokyo Institute of Technology, Interdisciplinary Graduate School of Science and Engineering, Professor, 大学院・総合理工学研究科, 教授 (60293073)
|
|---|
| Co-Investigator(Kenkyū-buntansha) |
YASUMURA Yoshiaki Kobe University, Faculty of Engineering, associate Professor, 工学部, 助教授 (30302962)
|
|---|
| Project Period (FY) |
2002 – 2004
|
| Keywords | ADR / mediation / retrieval of judicial precedents / animated agent / case base / Bayesian Network / decision tree |
|---|
| Research Abstract |
(1)Development of Retrieval Subsystem for Provisions, Old Mediation Cases, and Situations of Mediation :
We have developed three retrieval subsystems for Old Cases which use different technologies. Especially, Situation Retrieval system searches for similar situation from a case base of mediations by focusing on the flow of issue points. The retrieved data is used to navigate the mediation process.
(2)Development of Controlling Technologies of Animated Agents :
To develop a mediation agent which chairs the mediation, we have developed the controlling technologies of facial expression of animated agents by constructing a mental model based on a Bayesian Network. We showed that by controlling facial expression during mediation, we can conduct mediation smoothly.
(3)Implementation of Cyber ADR system :
We unified subsystems described above into a Cyber ADR system. This system helps the participants of Online Disputation by preparing useful user interface such as animated agents, a diagram showing a flow of issue points, searching for similar situations and so on.
(4)Application to Education of Mediators :
Using the Cyber ADR system, we gathered 11 mediation cases, and analyzed them statistically. We showed we can classify the mediation cases by clustering techniques, and we showed we can recognize the mediation skill of the mediator by observing the flow of issue points.
|
