Computing rich semantic models of text in legal domains
Wai Kiang Yeap · Information & Communications Technology Law · 1998
The richness provided in a deep semantic model of text is appealing and yet few such models have been developed. What are the problems with existing practical NLP systems and why is developing such a model so difficult? In this paper, I discuss some of these issues and argue that a possible solution must focus on the reasoning process using knowledge of words rather than the use of other mechanisms and especially those that speed up the so‐called pre‐processing stage. It is also suggested that computing representations of text that are transcripts of judges’ oral reports on Family Law cases is an interesting and challenging test area.