The Importance of Lexicalized Syntax Models for Natural Language Generation Tasks

Abstract

The parsing community has long recognized the importance of lexicalized models of syntax. By contrast, these models do not appear to have had an impact on the statistical NLG community. To prove their importance in NLG, we show that a lexicalized model of syntax improves the performance of a statistical text compression system, and show results that suggest it would also improve the performances of an MT application and a pure natural language generation system.

Um agente de jornalismo automatizado cobrindo a Amazônia Azul.

“…In many cases, this step can be done trivially by hard-coding specific words or phrases for each domain concept. In some cases, however, fluency can be improved by allowing the NLG system to vary the words used to express a concept or relation, either to achieve variety or to accommodate subtle pragmatic distinctions (DAUM É III et al, 2002). For our Blue Amazon agent, the system chooses a lexicalization template for each structured sentence.…”