第二字节编码范围为0x400x7E与0xA10xFE-NLPIR自然语言处理.ppt

Introduction to Computational Linguistics and Basic Tech. Related计算语言学导论与基础技术 Hua-Ping Zhang（张华平） 副研究员 研究生导师 北京理工大学 计算机语言信息处理研究所 kevinzhang@ Preface Why do we have the lecture? A lesson from Japanese GrapeCity. Another experience with Chinese Trade Network A successful product in Tsinghua TongFang Group I wish it could assist you to get the proposed solutions for most problems in natural language as soon as possible and accomplish you related work effectively. Preface II What could you learn from the lecture? What is computational linguistics(CL)? Understand key problems in CL and the appropriate solutions. Emphasize on Chinese information processing, especially with lexical analysis. Familiar with commonly-used information tech., such as retrieval, filtering, clustering and categorization Preface III How about the schedule? Introduction and basic tech. (character encoding system, electronic dictionary) 2 hours Key tech. in Chinese information processing (emphasizing on lexical analysis and system ICTCLAS, introducing parsing, semantics) 4 hours Overview on information tech. and some samples. (IR,IF,SE, Clustering, Categorization,Abstraction,Semantic computing) Preface IV What is the difference between the lecture and others? Concise introduction, not detail instruction but large coverage More on technology rather than theory More on experience than technology Any advice, inquiry and discussion is appreciated. Outline Introduction to computational linguistics Chinese Character encoding system Electronic dictionary structure and management 计算语言学定义 计算语言学是一门以计算为手段对自然语言进行研究和处理的科学。 Computational Linguistics Natural language processing Wherever there is Artificial Intelligence, there is Artificial Stupidity.“哪里有人工智能，哪里就有人工愚蠢”。 计算语言学的研究对象 计算语言学的研究对象是自然语言 自然语言与形式语言的本质区别 歧义性 自然语言是一种符号系统 语言符号的特点（索绪尔） 任意性：语言符号的选择是任意的 线条性：语言符号的排列是线性的 语言、思维与客观世界 语言的层面 语言研究的层面 语音 语法（包括词汇层和句法层） 语法研究要回答的问题是：一句话为什么可以这么说而不能那么说？ 语义 语义研究要回答的问题是：这句话说了什么？ 语用 语用研究要回答的问题是：为什么要说这句话？ 语言的层面III 语言各层面之间的