摘要： 根据林产品贸易文本信息推送中信息结构化存储的需要，结合语义识别的基本原理和基于规则的信息抽取方法，提出一种基于规则的林产品贸易文本信息抽取方法，利用林产品贸易文本信息的特征，定义林产品贸易文本信息的文本层次识别规则，采用创建数据库和数据表匹配识别规则，给出识别规则匹配的正则表达式和文本内容截取识别规则，以抽取需要的特定事实信息，并以一种结构化的形式存储于数据库中。通过对实际林产品贸易网站的文本信息结构化抽取，证明该研究在林产品贸易信息推送中具有较好的应用价值。

关键词: 语义, 林产品, 贸易文本信息, 结构化, 信息抽取, 识别规则

Abstract: Based on the needs of structured storage of information in the forest products trade text messages information push and combined with the basic principle of semantic recognition and the rule-based information extraction, a research on forest products trade text messages structuring based on semantic is proposed. Took advantage of the characteristics of forest products trade text messages, this paper defines the level of text recognition rules in the trade text messages, uses match identification rules of creating databases and data tables, defines the regular expressions with matching identification rules and the rules of intercept text recognition to extract the special factual information. The information is stored in the database as a structured form. Through the text structured information extraction in the trade text messages, it proves that the research has good value in the forest products trade information push.

Key words: semantic, forest products, trade text messages, structuring, information extraction, recognition rules

中图分类号: 

• TP391