Intelligent Language Processing

スポンサーリンク

概要

• 論文の詳細を見る

• Intelligent systems for natural language processing (NLP) must be based on knowledge of the language(s) they are to process and also knowledge about the real-world domain of discourse. All knowledge-based AI systems are subject to a tension between achieving computational efficiency and utilizing enough knowledge to perform intelligently. A two-pronged approach to this problem is useful for NLP systems. (1) Unification grammars represent linguistic knowledge in a computationally tractable way and so permit the development of NLP systems with ample knowledge of language. (2) NLP systems can be designed as cooperative agents to interact with users, who supply certain world knowledge not built into the systems. Useful applications exist for NLP with limited world knowledge, as well as with broad knowledge. If unification-like methods extend to handle world knowledge, much more intelligence can be put into practical systems. NLP is important for human-machine interfaces, groupware, and programming languages. Greatly expanded basic research on natural language and its processing is needed.

• 社団法人人工知能学会の論文
• 1991-01-01

著者

• Peters S

  Stanford Univ.

• Peters Stanley

  Stanford University

関連論文

• Intelligent Language Processing

スポンサーリンク