The Lexicon Acquisition Problem

Although in the end our goal is to acquire an entire natural language interface, we currently divide the task into two parts, the lexicon acquisition component and the parser acquisition component. In this section, we discuss the problem of acquiring semantic lexicons that assist parsing and the acquisition of parsers. The training input consists of natural language sentences paired with their meaning representations. From these pairs we extract a lexicon consisting of phrases paired with their meaning representations. Some training pairs were given in the previous section, and a sample lexicon is shown in Figure 2.

  

Figure 2: Sample Semantic Lexicon