Abstract
Case-Based Reasoning (CBR) means solving problems by utilizing the experience gathered in precedent cases. The legal domain is an ideal domain for CBR since it relies on extensive utilization of precedents. The research presented in this paper deals with constructing the prototype of a computerized working system called The Judge's Apprentice. This system is designed for practical use in Israeli criminal cases to aid sentencing in two categories of crimes. The proposed system is intended to reinforce the principle of uniformity in sentencing i.e. passing similar sentences for similar offenses. Senior legal experts have welcomed the system. Most of them were of the opinion that the system could definitely be of advantageous use to judges and that the idea of using CBR for sentencing is relevant and should be applied. The system has been tested on 54 non-representative cases. A rather high success rate has been achieved in cases for which a precedent with high similarity has been retrieved. Future directions for research are: elaboration of the model for the handling of other types of offenses; elaboration of the model to other domains requiring case-based evaluation.
| Original language | English |
|---|---|
| Pages (from-to) | 191-202 |
| Number of pages | 12 |
| Journal | New Review of Applied Expert Systems and Emerging Technologies |
| Volume | 5 |
| State | Published - 1999 |
