Closing the loop in webpage understanding
Proceedings of the 17th ACM conference on Information and knowledge management（2008）
摘要
Little work has been done towards an integrated statistical model for understanding webpage structures and processing natural language sentences within the HTML elements. This paper proposed a novel framework called WebNLP which enables bidirectional integration of page structure understanding and text understanding in an iterative manner. Experiments show that the WebNLP framework achieved significantly better performance.
更多查看译文
关键词
conditional random fields,html,natural language processing,natural languages,html elements,internet,helium,information extraction,top down,data mining,markov processes,natural language,labeling,text segmentation,text analysis,knowledge management,conditional random field,statistical model
AI 理解论文
溯源树
样例
生成溯源树，研究论文发展脉络
Chat Paper
正在生成论文摘要