ABSTRACT
This chapter discusses the semantic tools that let us go beyond the capabilities of syntax-based tutors such as BRIDGE as well as the ALICE-chan and CALLE systems. The central idea of Lexical Conceptual Structure (LCS) is that the human language can be modeled using a uniform internal representation of conceptual information with a few parameters that can be toggled one way or another to obtain various languages. The versatility of the LCS allows us to reuse labor-intensive features of the lexicon across languages. The versatility of the LCS allows us to reuse labor-intensive features of the lexicon across languages. The parse tree produced by the syntactic analyzer is passed to the semantic analysis or composition component, and the language-independent LCS representation is generated. This representation is then passed to the LCS matcher, which checks the answer for a match against the desired answer that was prestored by the author.
