Relation extraction based on multi-view ensemble algorithm

Learning from labeled data is well-established in machine learning, whereas building a large amount labeled data set is time consuming and often expensive. Therefore, more and more researchers pay attention to unsupervised machine learning. This paper focuses on unsupervised machine learning, and a multi-view ensemble learning algorithm is proposed for relation extraction. First, we design three sufficient and conditionally independent views to capture different attributes parts of the data, and build classifiers based on three views. Then, every two of the three classifiers are used to label some unlabeled instances for the third one. Finally, we propose to generate multiple base models using traditional machine learning algorithms and combine them into an ensemble model. Different feature spaces and different machine learning algorithms can be combined to improve the performance of the relation extraction by using multi-view ensemble learning. And this multi-view ensemble based relation extraction system is under development. ICIC International