网络信息提取系统研究

A Survey of Web Information Extraction Systems

阐述网络信息提取系统的研究现状。从信息提取技术和自动化程度两方面对现有网络信息提取系统进行对比,由此将网络信息提取系统分为非自动化、半自动化和全自动化三类。综合考虑标记方法、提取规则类型和特征、学习算法、用户参与度、适用性以及输出接口等因素,对三类系统的性能优劣进行评估。最后对网络信息提取系统进一步的研究工作进行了展望。

The related work of Web information extraction systems is introduced. Comparison is done in two dimensions： extraction technology and automation degree. Web information extraction systems are divided into three types： non-automated, semi-automated and fully automated. The performance of three types of systems is evaluated according to the comprehensive considerations on some fac- tors, such as tokenization method, the types and features of extraction rules, learning algorithms, user participation, fitness and output interface. Finally. the further research work is nrosnected.