CLIN 2005 Abstracts
  • Inducing Head-Driven PCFGs with Semantic Heads
    Detlef Prescher (Universiteit van Amsterdam)
    Although state-of-the-art parsers for natural language are lexicalized, it was recently shown that accurate unlexicalized parsers for the Penn tree-bank can be simply read off a manually refined tree-bank. While lexicalized parsers often suffer from sparse data, manual mark-up is costly and largely based on individual linguistic intuition. Thus, across domains, languages, and tree-bank annotations, a fundamental question arises: Is it possible to automatically induce an accurate parser from a tree-bank without resorting to full lexicalization? We show how to induce head-driven probabilistic parsers with head classes from a tree-bank, and we demonstrate that our automatically trained parser has a performance already better than that of early lexicalized ones. Furthermore, we exemplify that the head classes are semantic in nature.