中文文本中时间信息抽取及规范化方法

Temporal Information Extraction and Normalization Method in Chinese Texts

时间信息抽取和规范化表达在时空泛在信息处理、关联、聚合和内容服务中具有重要作用。设计并实现了一种基于规则的时间信息抽取和规范化表达的方法。首先,结合中文文本中时间信息描述的特点,定义了基本时间单元和派生时间单元,并由时间单元组合形成不同的时间表达式;然后,构建了时间词汇词典和时间表达式抽取的正则表达式模板;最后,设计了时间推理和计算的方法,实现了时间信息的规范化表达。基于中文文本语料进行了时间信息抽取测试。实验结果表明,本文方法具有较好的时间信息抽取效果。

Temporal information extraction and normalization expression play a significant role in information processing, association, aggregation and content service. In this paper, a method of temporal information extraction and normalization expression based on a regulation are designed and realized. First of all, combining the characteristics of temporal information description in Chinese text, the basic time unit and derived time unit are defined. Secondly, a regular expression model of temporal vocabulary glossary and the extraction of time expression is built. Finally, a method of reasoning and calculating time is designed, and the normalized expression of temporal information is realized. Then, a test on the extraction of temporal information based on Chinese text corpus is carried out, the result shows that this method has a good effect on temporal information extraction.