Weakly-Supervised Semantic Relation Extraction Using Stratified Strategy
Pan Shen · Journal of Guangxi Normal University · 2008
This paper presents a novel method of using bootstrapping for the task of weakly-supervised semantic relation extraction.The initialized training set are chosen based on the stratified strategy.Experiments show that the initialized training set play an important role in weakly-supervised semantic relation extraction,and the proposed method outperforms that of random selection.