节点文献
中文信息处理研究工作的新进展
Recent Advances of Chinese Information Processing Research
【摘要】 中文信息处理研究应以应用开发带动理论研究和技术研发。我国中文信息处理研究已在三个方面取得了重要进展:应用理论研究方面,语义的知识表示方法从句子扩展到了句群层面,构建了概念—语句—句群统一的知识表示模型;知识库资源建设方面,建立了达到应用规模的、包括概念知识库、语言知识库和言语知识库的多层次语义知识库资源;中文智能搜索引擎核心技术研发方面,实现了文本的语义分析和多层次的语义索引技术,以及基于语义理解的文本相似度计算技术,使用户可以用最自然的语言提出搜索请求,而不必受关键词的约束。基于语义的新一代中文搜索引擎即将出现大的突破。要提高我国中文信息处理的竞争力,应当重视具有自主知识产权的基于语义的理论和应用研究,加强人才队伍建设,进行科学组织和管理,促进以智能搜索技术发展及应用为主线的中文信息处理产业联盟的构建。
【Abstract】 Driven by the demand of application,Chinese information processing research has made great progress in the aspects of semantic theory,knowledge base construction,semantic analysis and their applications in the search engine. In the theoretic research,the semantic representation has extended from the sentence level to the sentence group level,and a unified representation model based on the concepts has been established. A large semantic knowledge base has been built,which includes concept knowledge,linguistic knowledge and world knowledge,and can meet the need of future development. As for the search engine,the semantic analysis has been used to analyze the Chinese text and extract situation information from the context,the strategy of multilevel index has been designed,and the algorithm of relevance computing based on semantic understanding has been used. These researches can give powerful support for the intelligent search engine to deal with the user’s requirement for a natural language. To enhance the competitiveness of Chinese information processing technology,the semantic research with completely independent intellectual property rights should be strengthened,more talents should be trained,and the industry alliance should be formed to promote the application of Chinese information processing technology in a scientific way,especially in terms of the intelligent search engine.
【Key words】 Chinese information processing; applied theory; knowledge base; search engine;
- 【文献出处】 云南师范大学学报(哲学社会科学版) ,Journal of Yunnan Normal University(Humanities and Social Sciences) , 编辑部邮箱 ,2010年04期
- 【分类号】TP391.1
- 【被引频次】10
- 【下载频次】793