Case based reasoning in legal case-retrieval

K. Ashley (1992) has proposed that, instead of building legal expert systems that answer legal problems, it may be better to build systems that assist lawyers in finding source material with which they may answer their own problems. Legal database systems should make information available tailored to the particular context of the user's problem and, in doing so, would use the same relevance criteria as human experts. Two case-based reasoning systems were developed (in CA-Clipper) which attempt to fulfil Ashley's criteria: CAMLAW (CAseMark LAW), for the retrieval of industrial law reports, and ICARUS (Intelligent CAse Retrieval Using Semantic Comparison). CAMLAW could be regarded as fulfilling Ashley's criteria. ICARUS, at its current stage of development has some problems. Something better than Roget's Thesaurus needs to be explored. Longman's Dictionary of Contemporary English, a machine-readable dictionary might be a better choice. Then the linkage problem needs to be examined in greater detail.