Computational Models of Referring

Abstract

This chapter discusses computational models of the act of referring, focusing on limitations of the classic algorithms in this area, as illustrated by a number of common referential scenarios. After a brief introduction to the computational enterprise known as Referring Expressions Generation, the chapter focuses on a range of issues that can arise from the situation in which a referring expression is uttered. The chapter discusses situations in which there is uncertainty over the hearer’s domain knowledge, situations in which the referent cannot be pinned down precisely, and situations in which identifying the referent is not the main goal. It is shown that situations of this kind complicate the referential task, and strategies that speakers use to cope with them are discussed.