基于有限状态自动机提取不规范表结构Web信息

Unregulated table structure Web information extraction based on finite state automata

大量的不规范表结构信息是当前Web信息提取所必须解决的问题.在现有方法基础上,给出了归纳学习相邻属性间上下文规则集算法,提出了以Web页为粒度的属性转换机和有限状态自动机包装器概念,最后介绍了采用有限状态自动机包装器提取不规范表结构Web信息的算法.

lots of unregulated table structure information currently come to be the unavoidable issue of Web information extraction. Based on the existing method, a further research about inducing contextual rules of adjoining attributes has been done; and a new conception of the finite state automata wrapper and attributes transducer, whose granularity is Web pages, is presented. Finally, the algorithms for the unregulated table structure Web information extraction by finite state automata wrapper, are introduced.