为了解决中文文本中时间信息的形式表达、抽取及时间语义计算等一系列问题,提出了一种可以适用于多种自然语言处理任务的时间信息处理系统,通过对汉语句子时间要素的分析以及时间词语构成形式的研究,采用一种时间表达式的概念,将承载时间信息的短语按照功能的不同分解成若干容易识别,语义单一的“小”的成分,并在此基础上给出基于规则的时间信息抽取、理解及时间语义的计算方法.
In order to resolve a series of problems including representation, extraction and temporal semantic computation of Chinese text temporal information, a temporal information processing system which can be applicable to various natural languages processing mission is proposed. After analysing Chinese sentences temporal elements and component of temporal phrase, adopting a concept of temporal expression, the phrase expressing temporal information were separated into several sub-constituents which have unitary semantic and can be identified easily. On this basis, the methods of extracting, comprehending and semantic computation of temporal information based on the rule were studied.