Learning Information Extraction Rules for Semi-Structured and Free Text
Machine Learning - Special issue on natural language learning
Snowball: extracting relations from large plain-text collections
DL '00 Proceedings of the fifth ACM conference on Digital libraries
Extracting Patterns and Relations from the World Wide Web
WebDB '98 Selected papers from the International Workshop on The World Wide Web and Databases
Hierarchical hidden Markov models for information extraction
IJCAI'03 Proceedings of the 18th international joint conference on Artificial intelligence
Efficiently inducing features of conditional random fields
UAI'03 Proceedings of the Nineteenth conference on Uncertainty in Artificial Intelligence
Hi-index | 0.00 |
Quantities of valuable relation knowledge are contained in textual documents on the World Wide Web. However, those data are always organized in semi-structured text and cannot be used directly. We develop an automatic and effective approach to extract relations from World Wide Web, which just requires a few user specified seed instances as input. Those instances are used to generate extraction rules that in turn result in new instances. And in order to improve the reliability of results, an effective method is proposed to assess new extracted instances. This paper introduces the approach in details and the experimental results show that the approach achieves an average precision of 98.67% and can preferably complete the relation extraction task.