Robust semantic construction
Michael Schiehlen · 2000
Recent years have seen a surge in interest for robust flat analysis, i.e. NLP systems with fairly limited supply of linguistic knowledge but with vast coverage. The paper describes a module that serves as a back-end to such flat analysis methods and transforms their output into full semantic representations as constructed by deep analysis methods. In particular, the module has been designed so as to process input from -- tree banks -- a statistic context-free parser trained on these tree banks -- a finite-state parser -- a traditional feature-structure parser The semantic representations which the module constructs are so-called Verbmobil Interface Terms (VITs) (Bos et al., 1998), (building on Reyle's Underspecified Discourse Representation Structures (1993), see an example in Figure 1). Although in principle other representations could be constructed as well, VITs seem to be a...