Hypothesis selection in grammar acquisition
Masaki Kiyono, Jun’ichi Tsujii · 1994
This paper presents some techniques for selecting linguistically adequate hypotheses of new grammatical knowledge to be used as resources of grammatical knowledge acquisition. In our framework of linguistic knowledge acquisition, a rule-based hypothesis generator is invoked in case of parsing failures and all the possible hypotheses of new grammar rules or lexical entries are generated from partial parsing results. Although each hypothesis could recover the defects of the existing grammar, the greater part of hypotheses are linguistically unnatural. The techniques we propose here prevent such unnatural hypotheses from being generated without discarding plausible ones and make the following corpus-based acquisition process more efficient and more reliable.