中文信息处理研究现状分析.PDFVIP

中文信息处理研究现状分析* 宗成庆 （中国科学 院 自动化研究所 北京　 100190 ） 提 要 60 多年来 中文信息处理研究取得 了令人 瞩 目的成就 。但是 ，这一领域也 面 临 问题 和挑 战。本 文在对 中文信息 处理研究成就简要归纳 的基础上 ，分析这一领域 的技术现状 ，直面存在 的问题 ，并对未来发展 的方 向提 出一些看法 。希 望本文指 出的问题能够引起 中国国内同行 的关注 ，为未来 的中文信息处理研究提供有益 的参考 。 关键词 中文信息处理 ； 自然语 言处理 ； 自然语 言理解 ；计算语 言学 Chinese Language Processing: Achievements and Problems Zong Chengqing Abstract In the past over 60 years, research on Chinese language processing has made great achievements. With the rapid develop- ment and popularization of the Internet and communication technology, Chinese language processing technology has attracted world- wide attention in recent years. This article summarizes the achievements of Chinese language processing and analyzes the present status of the technology in this fi eld, particularly the problems that the fi eld may face in term of development. The author argues that it is still diffi cult for artifi cial intelligence to “understand” rather than “process” naturally produced Chinese because of the following three rea- sons: (1) the current information processing technology is inadequate in processing grammatically complex Chinese sentences; (2) there are unsolved problems in machine learning technologies; and (3) our understanding of how human brain processes language is still very limited. This paper concludes that we need a better understanding of how the Chinese language is decoded in human brain and build a computational model that specifi cally targets at the Chinese language in order for artifi cial intelligence to understand naturally produced Chinese. Key words Chinese language processing; natural language processing; natural language understanding; computational ling