Corpus-driven learning of Event Recognition Rules
msra(2000)
摘要
In this paper a complex framework for adaptaing IE systems to changing domains and users is described. The proposed methodology is based on the integration of different learni ng meth- ods over a corpus (i.e. example-driven conceptual clusteri ng, corpus- driven probabilistic learning and terminological reasoni ng) and on an available general-purpose ontology. First experiments have been carried out on domain-specific corpora (the annotated porti on of the PennTree Bank and the Reuters TREVI collection2) and used Word- net (16) as the reference ontology. However, the methodology is in- dependent from the specific domain as well as from the adopted on- tology. Early evaluation is promising and first results will be pre- sented and discussed.
更多查看译文
AI 理解论文
溯源树
样例
生成溯源树,研究论文发展脉络