Prior Knowledge

In order to facilitate the learning process, we introduce prior knowledge into the system. This knowledge provides a formal and general description of the linguistic structures that ALLiS can learn. The essential kind of knowledge we introduce consists on more appropriate and distributional categories than the ones previously used. We show that if a larger set of categories is used and if these categories are more suitable, the learner's task is eased (the processing time is reduced, as well as the number of rules) and the quality of the rules is better. The following sections describe in detail the different types of prior knowledge we use.