In recent years there has been a resurgence of interest in empirical corpusbased approaches to natural language processing driven by the belief that empirical statistical methods can succeed at many of the problems where rationalist techniques have failed.
